Fetal Fibronectin Testing for the Diagnosis of Preterm
Labour: A Review of the Economics Literature and
Cost Analysis for Alberta
Report prepared for the Alberta Health Technologies
Decision Process: Fetal Fibronectin Project
Gillian R. Currie, PhD ([email protected])
Assistant Professor
Department of Paediatrics
Department of Community Health Sciences
University of Calgary
Research Fellow, Institute of Health Economics
June 2006
Acknowledgments:
The author would like to thank the following for their assistance with various aspects of
the preparation of this report:
•
•
•
•
•
•
Diane Lorenzetti for assistance with literature search,
Seija Kromm for research assistance with the literature review,
Helen Lee for assistance with data analysis,
Margaret Wanke, Charis Management Consulting, Inc., Project Manager for the
Fetal Fibronectin Project,
Fetal Fibronectin (fFN) Project Team,
Fetal Fibronectin (fFN) Expert advisory Group.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
© 2006, Institute of Health Economics
Table of Contents
1.0
BACKGROUND ............................................................................................................................ 5
2.0
OBJECTIVES ................................................................................................................................ 6
PART I: THE ECONOMICS OF FETAL FIBRONECTIN TESTING FOR THE DIAGNOSIS OF
PRETERM LABOUR: A LITERATURE REVIEW ................................................................................ 8
3.0
INTRODUCTION.......................................................................................................................... 8
4.0
METHODS ..................................................................................................................................... 8
5.0
RESULTS OF THE LITERATURE REVIEW........................................................................... 9
5.1
5.2
5.3
6.0
NON-EXPERIMENTAL EVIDENCE .................................................................................................. 9
DECISION MODELLING APPROACH .............................................................................................. 12
EXPERIMENTAL STUDY DESIGN .................................................................................................. 14
CONCLUSIONS .......................................................................................................................... 18
PART II: ANALYSIS OF COST IMPLICATIONS OF FFN TESTING FOR ALBERTA................ 20
7.0
INTRODUCTION........................................................................................................................ 20
8.0
DATA EXTRACTION STRATEGY FROM ALBERTA HEALTH DATABASES ............. 20
9.0
POPULATION CHARACTERISTICS OF WOMEN PRESENTING WITH
THREATENED PRETERM LABOUR ................................................................................................... 21
9.1
9.2
9.3
9.4
9.5
10.0
NUMBER OF WOMEN ................................................................................................................... 23
GESTATIONAL AGE AT BIRTH ..................................................................................................... 24
REGIONAL DISTRIBUTION ........................................................................................................... 26
AGE STRUCTURE ........................................................................................................................ 28
SOCIOECONOMIC STRUCTURE .................................................................................................... 29
COST IMPLICATIONS.............................................................................................................. 31
10.1
COSTS OF FFN TESTING .............................................................................................................. 31
10.2
POTENTIAL NUMBER OF TESTS ................................................................................................... 32
10.3
VARIABLE COSTS OF FFN TESTING ............................................................................................. 33
10.4
FIXED COSTS .............................................................................................................................. 34
10.5
LENGTH OF STAY ....................................................................................................................... 34
10.6
AVERAGE COSTS FOR THREATENED PTL ................................................................................... 36
9.6.1
Inpatient admission............................................................................................................... 36
9.6.2
Physician billing ................................................................................................................... 36
9.6.3
Potential costs avoided from LOS reductions....................................................................... 36
10.7
AMBULANCE TRANSFERS ........................................................................................................... 37
10.7.1
Numbers of Transfers between facilities .......................................................................... 37
10.7.2
TOTAL AMBULANCE TRANSFERS ........................................................................................... 39
10.7.3
Costs of ambulance transfers........................................................................................... 40
10.7.4
Cost savings from avoided ambulance transfers.............................................................. 40
11.0
OVERALL COST IMPLICATIONS ......................................................................................... 40
12.0
CONCLUSIONS .......................................................................................................................... 43
DISCUSSION: ..............................................................................ERROR! BOOKMARK NOT DEFINED.
REFERENCES ........................................................................................................................................... 45
APPENDIX 1 .............................................................................................................................................. 47
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
© 2006, Institute of Health Economics
APPENDIX 2 .............................................................................................................................................. 53
APPENDIX 3 .............................................................................................................................................. 63
APPENDIX 4 .............................................................................................................................................. 66
APPENDIX 5 .............................................................................................................................................. 69
APPENDIX 6 .............................................................................................................................................. 72
APPENDIX 6 .............................................................................................................................................. 72
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
© 2006, Institute of Health Economics
Fetal fibronectin testing for the diagnosis of pre-term labour
1.0
Background
Preterm birth (PTB) is one of the biggest problems in modern obstetrics. Preterm birth is
responsible for the majority of perinatal mortality and adverse long-term neurological
outcomes. Preterm birth rates in Alberta are higher than the national rate, and both have
been rising over time. In 2004, in Alberta almost 9% of live births were preterm. PTB
can arise due to spontaneous labour with or without premature rupture of membranes, or
it can be as a result of an intervention induced by fetal or maternal complications. For
spontaneous preterm labour (PTL) without rupture of membranes, the majority of women
presenting with symptoms will go on to deliver at term (over 80% deliver at term). Thus,
the challenge in practise is to correctly identify those women who are in true preterm
labour and will deliver preterm.
The objective of early diagnosis is transfer to an appropriate tertiary centre, administer
corticosteroids to help maturation of immature neonatal lungs, perhaps administer
antibiotics and possibly administer tocolytics to prolong pregnancy. Correct diagnosis
means identifying which patients need these treatments and which do not. The fetal
fibronectin (fFN) test is a diagnostic test that can be used to help diagnose PTL in women
with signs and symptoms of labour, intact membranes and less than 3 cm cervical
dilation. The signs and symptoms of preterm labour are regular uterine contractions, low
abdominal cramping, low back pain, pelvic pressure or increased vaginal discharge. The
positive predictive value of the test for delivery within 7 days is 20%, meaning that 20%
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
5
© 2006, Institute of Health Economics
of women testing positive will go on to deliver within 7 days. The negative predictive
value of the test for delivery within 7 days is 99%, meaning that 99% of women testing
negative will not deliver within 7 days (Norman and Greer, 2005).
The primary advantage of the test is its simplicity of use and the strong negative
predictive value. This potentially gives the ability to more accurately identify those
women who are in true preterm labour, and who therefore need to be transferred and
treated. Importantly, transfer and treatment can be avoided for those women determined
not to be in true labour. This can save health care resources in transferring patients and in
inpatient admissions for preterm labour, as well as provide peace of mind and enable
women to remain in their community and homes.
2.0
Objectives
This report was prepared as part of the Alberta Health Technologies Decision Process
review of fFN testing in Alberta. In addition to this report, the AFHMR Health
Technology Assessment Unit prepared a report entitled “The role of the rapid fetal
fibronectin assay in the management of suspected preterm labour” which contains details
on the biological basis of the test, the mechanism of the test, and on diagnosis and
management of PTL including the role of fFN, the regulatory status in Canada,
unpublished Canadian studies of the impact of fFN (including the economic impact) and
reviews published systematic reviews and health technology assessment studies.
Therefore, these topics are not duplicated in this report.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
6
© 2006, Institute of Health Economics
The objectives of this report are as follows:
1. Review of the published literature examining economic impact of fFN test.
2.
Describe population characteristics of Alberta women presenting with symptoms
of PTL (number, age, socioeconomics, regional distribution).
3. Examine the cost implications of fetal fibronectin testing in Alberta.
Part I of this report contains the review and summary of the published literature on the
economic impact of fFN testing. Part II of this report addresses the second and third
objectives using administrative data.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
7
© 2006, Institute of Health Economics
Part I: The Economics of Fetal Fibronectin Testing for the
Diagnosis of Preterm Labour: A Literature Review
3.0
Introduction
The purpose of Part I of this report is to summarize the available published evidence on
the economic impact of fetal fibronectin testing for the diagnosis of preterm labour. As
noted previously, there is also additional Canadian evidence which does not yet appear in
the published literature. That unpublished evidence is not presented in this document, but
is summarized in the AFHMR report prepared as part of the Fetal Fibronectin Health
Technology project.
4.0
Methods
A review of the literature was conducted to identify relevant articles for this review.
Relevant databases were searched as well as bibliographies of the selected papers. We
selected studies for review which examined the impact of fetal fibronectin testing on
hospital admission, length of stay, maternal transfers, use of tocolytics and corticosteroids
as part of a comparative study of different strategies for the management of preterm
labour, including fFN or not. The impact on quality of life, including maternal anxiety
and stress was also considered. Details of the search strategy can be found in Appendix
1.
In the following section, the results of the literature review are presented, and the
evidence available from studies using non-experimental design approaches, including
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
8
© 2006, Institute of Health Economics
decision modelling, and experimental (randomized controlled trial) designs is presented
and critically reviewed.
5.0
Results of the Literature Review
Ten papers were selected for full review. The papers selected are identified and
summarized in a Table in Appendix 2. The published literature on the economics of fetal
fibronectin testing ranges from 1999 through 2005 and includes both non-experimental
and experimental study designs (cohort studies pre- and post-introduction of fFN and also
randomized controlled trials) and two decision-analytic modelling studies. The evidence
comes from primarily out of the United States (seven studies), with a single study from
each of Canada, New Zealand and Australia. Most of the published studies look at the
impact of fFN testing at tertiary care hospitals only, with only two studies that considered
referring hospitals as well (Giles et al, 2000, Joffe et al, 1999). Only the Giles et al study
however, explicitly considered the effect on transfers; the Joffe et al study did not
identify transfers but noted that patients were selected who presented to either physician’s
offices or the hospital.
5.1
Non-experimental evidence
The conclusion based on the non-experimental evidence is that the potential for changing
the management of preterm labour and reducing admissions for threatened preterm labour
promised by the high negative predictive value of fFN testing can be realized in practice.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
9
© 2006, Institute of Health Economics
The first studies came out of the U.S. and Australia (Joffe et al 1999, Giles et al 2000).
Joffe et al 1999 was a pre-post fFN comparison of admissions, and tocolytic prescriptions
as well as neonatal outcomes. Compared to baseline, there were significantly fewer
admissions for PTL, fewer prescriptions written, fewer admissions per patient and shorter
length of stay. There were no differences in neonatal outcomes (NICU admissions,
length of stay, and days of ventilator support). This was one of the few studies which
considered neonatal outcomes.
The Giles et al 2000 study was based upon an 18 month prospective audit of fFN use, and
did not actually compare the fFN tested group with an untested historical control group –
thus its estimates of cost savings are misleading. They compare the management of the
fFN positive group and the fFN negative group and estimate costs savings of fFN for
transfers based on an assumption that in the absence of fFN testing all patients who tested
negative would have been transferred. This would overestimate the costs savings. This
study was the only one that explicitly looked at transfers from rural hospitals to tertiary
facilities, thus it is unfortunate that the study is flawed and renders the conclusions about
transfer uncertain.
The only published Canadian evidence comes from Abenhaim et al 2005 which looked at
the impact of fFN testing on hospital resource utilization at a university-associated
tertiary care hospital in Montreal. A prospective cohort of 116 women presenting with
preterm labour after fFN testing was available was compared with an historical cohort of
women prior to the availability of fFN. There is little information given about the
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
10
© 2006, Institute of Health Economics
selection of the historical cohort other than to say they were identified through a
systematic review of all birthing centre triage visits, admissions and discharges – as was
the cohort for the post-fFN period. Overall, in the post-fFN time period, there were lower
admissions for PTL with eventual discharge undelivered, and there was a statistically
significant 88% reduction in days hospitalized and lower total cost. The post-fFN study
period did not begin until six months after its introduction to allow a learning period by
clinicians in terms of the use and interpretation of the test. It is important to note that not
all patients in the post-fFN period actually had an fFN test conducted – including some
who were admitted and delivered preterm, and others who were discharged with alternate
diagnoses. 47 fFN tests were done (out of 116 women) – the rest were diagnosed based
on clinical criteria.
If fFN tests had been done on all women, this may have impacted
the results – highlighting the importance of an explicit fFN protocol to guide how the test
can be incorporated into the management of PTL. It is also important to note that 25% of
the fFN tests conducted were “inappropriate” (although half of these were for twin
gestations, which other studies did not exclude) - these costs were included in the
analyses, so again this highlights the importance of education around the appropriate use
of the test as well as potentially greater gains to be made in cost savings.
The most recent study adopting a non-experimental design was conducted in New
Zealand (Mussaad et al 2005). This is the only of these studies not to find a difference in
spending on management for PTL. They did find a difference in the use of tocolytics and
steroids, however this did not translate into cost savings. There was no difference in
length of hospital stay. They also found no difference in neonatal outcomes (NICU
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
11
© 2006, Institute of Health Economics
admissions, and invasive ventilation). However, this was a very small study with only 30
patients in the post-fFN group matched with 30 historical controls and there are issues
around patient selection. They also considered the first 30 fFN tested patients so did not
allow any time for a settling in period for familiarization with the test and how to
interpret it. They specify that all patients testing positive are admitted which may not be
an appropriate management strategy based on the poor positive predictive value of the
fFN test. For all these reasons, the results of this study are not particularly robust.
5.2
Decision modelling approach
There were two studies which adopted a decision modelling approach to examine the
potential impact of fFN testing. Decision models create a set of strategies which might
be adopted, incorporate data from local sources as well as typically relying in large part
on data from the literature, and examine the relative costs and effects of those different
strategies.
The first published decision model was the Mozurkewich et al 2000 study.
They
consider six different strategies for the management of threatened pre-term labour
(patient presents with regular contractions, between 24-34 weeks, with intact membranes
and cervical dilation <3 cm). The strategies considered are (1) treat all with tocolytics
and corticosteroids, (2) administer ‘traditional’ fFN test and in meantime treat all,
discontinuing those who are negative, (3) administer cervical ultrasound (transvaginal or
transperineal) and treat only with abnormal cervical length measurements, (4) administer
‘rapid’ fFN and treat those who are positive, (5) administer rapid fFN plus ultrasound and
treat those with abnormal results on at least one, (6) treat none. They further consider an
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
12
© 2006, Institute of Health Economics
additional 3 strategies – (7) treat all women with outpatient corticosteroids only, (8)
corticosteroids to all, tocolytics only to those with positive rapid fFN test, and (9)
corticosteroids to all, tocolytics only to those with abnormal cervical length.
The
majority of the data for the decision model was taken from the literature, and also from
local institutional data (Michigan).
The outcome focus was on respiratory distress
syndrome (RDS) and neonatal death. They found that the cheapest strategy was the treat
all with corticosteroids, the strategy with the lowest number of cases of RDS was the
cervical length plus corticosteroids strategy, and the strategy with the fewest neonatal
deaths was the same for treat all as cervical length plus corticosteroids (and almost the
same for both fFN strategies).
Compared with treating all women with threatened
preterm labour, the authors conclude that risk assessment strategies using either fFN or
cervical length assessment offer cost savings and prevent cases of respiratory distress
syndrome and neonatal death. The addition of corticosteroids improves the clinical
outcomes at lower costs.
The Sullivan et al 2001 adopted a decision-modelling approach to assess the question of
the impact of fFN testing in reducing costs of managing PTL. They use local data on PTL
from a U.S. hospital and data from the literature on fFN testing as estimates in their
model. They compare three strategies fFN testing on all, examination alone and fFN
testing only on those for whom clinical criteria suggest admission. They find that costs
can be higher if fFN is done on all women, rather than on those for whom clinical criteria
alone can not rule out PTL. The admission rates rise in the presence of fFN testing – they
assume that all positive fFN patients would be admitted. Some of the estimates built into
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
13
© 2006, Institute of Health Economics
the model are not based on any available evidence - in particular, the assumptions about
the prevalence of a positive fFN assay in the context of the different strategies.
5.3
Experimental study design
The strongest study design to answer the question of the impact of fFN testing is an
experimental design - that is, randomized controlled study design. There are four reported
randomized controlled trials, all conducted in the U.S. at tertiary care hospitals: three
have been published (Grobman et al 2004, Lowe et al 2004, Plaut et al, 2003) and an
additional one is published in abstract form (Nguyen et al 2002).
The findings from
these studies are mixed. They do not, in general, provide strong evidence that fFN testing
reduces health care utilization:
the Grobman et al, 2004 study of 100 randomized
patients concludes that fFN testing did not reduce the health resource utilization, the
Lowe et al, 2004 study of 100 randomized patients found no significant difference
between groups in terms of administration of drugs, length of stay in labour and delivery,
admissions and length of stay in the antepartum ward using standard analytic techniques
but do find evidence of a shorter length of stay using a hazard model approach, and the
Nguyen et al 2002 study of 77 randomized patients conclude that fFN testing increased
time spent in triage and hospital charges. The Plaut et al, 2003 study randomized 100
patients and found no difference in length of stay in general, however they did note a
40% decrease in hospital stay for those women who were observed longer than six hours.
The Nguyen et al 2002 study has not been published in full, therefore it is not possible to
fully assess the adequacy of this evidence. There is insufficient evidence provided about
the study design, including the management of patients in the no fFN group. They
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
14
© 2006, Institute of Health Economics
indicate that all patients with a positive fFN result underwent further observation. This
appears to be an inappropriate use of the test. In addition, while the triage times were
statistically significantly different but it is not clear that an 18 minute difference is
meaningful from a policy perspective.
The Plaut et al 2003 RCT originally planned to focus on transfers to tertiary care
hospitals, but stopped their study early due to lack of enrolment. No reasons are provided
for the difficulty in enrolling patients – fFN testing was not available outside the study
protocol. The authors report an analysis of length of stay. While overall length of stay
was not significantly different between the group who had fFN test results available and
the group that did not, there was a 40% decrease in hospital stay for the group at higher
risk of PTL (i.e., those who had longer periods of observation). No fFN protocol was
developed and implemented as part of the trial, but the authors argue that they
“…attempted to mimic the real world, where physicians are educated about a test and
then try to incorporate it in their practice.”
The Grobman et al 2004 study focussed on whether fFN testing changed physician
behaviour and resultant health care costs. The fFN test had not been used in their
hospital setting prior to the study, and was only available within study protocol during the
study. The fFN swab was conducted following cervical exam for all patients who met the
inclusion/exclusion criteria but then patients were randomized into two groups: one
where the attending physician had the results available and one where they did not (the
swabs were frozen and analyzed as a batch later). Since the attending physicians were not
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
15
© 2006, Institute of Health Economics
familiar with using fFN in practice, an educational intervention was provided as well
alongside the trial: prior to the study a letter explaining the characteristics of the test was
sent to admitting obstetricians. Also when fFN test results were provided to physicians,
they were reminded of the meaning of the results with the delivery of a standardized
reminder.
No evaluation of the effectiveness of the educational intervention was
conducted.
The Grobman et al study found no difference in time spent in labour and delivery for
initial evaluation, frequency of tocolysis or corticosteroids, admissions or readmission
post-discharge. Sub-group analysis was conducted by time to assess whether there was a
learning curve in terms of altering behaviour – they found no difference, however the
study was not powered for this analysis.
This study fails to find a difference in
behaviour. It is unclear whether these admitted obstetricians are already doing a very
good job on correctly identifying PTL and thus there was not room for much
improvement or alternatively whether the educational intervention failed to adequately
inform clinicians how to use the fFN test results in supplementing their clinical
judgement. The authors conclude that further evidence is need and “perhaps, for example,
a standardized clinical pathway may need to be implemented if the value of these tools is
to be realized optimally.”
The Lowe et al 2004 study looked at the impact of fFN testing on length of hospital stay
and preterm labour interventions in the context of a tertiary care hospital in the U.S.
Their inclusion criteria allowed twin gestations and included greater cervical dilation than
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
16
© 2006, Institute of Health Economics
other studies for multiparous women. They included women who had already been
started on tocolytics or other therapies prior to transfer to the tertiary care facility in their
study, but do not report on the percentages in each arm, so it is difficult to interpret their
results on drug use. Overall, the results of this RCT did not find any difference in the two
groups in terms of time spent in labour and delivery, admission to the antepartum unit,
length of stay on the antepartum ward, use of tocolysis, steroids and antibiotics.
However, using Cox proportional hazards model, and adjusting for previous preterm
birth, cervical dilation and gestational age, they did find that fFN women had a shorter
stay in labour and delivery.
The authors did not report on any testing that the
assumptions of that model were met, however, it is of note that this study found this
difference using a different method of analysis than the standard unadjusted difference in
means between two groups. This raises the question of why these adjustments were not
made in other studies, as well, and what the implications may be for their conclusions.
In summary, the evidence from RCTs is mixed. Generally, little evidence supporting the
potential of fFN to change management of PTL, and reduce unnecessary treatment, was
provided. However, two studies, Plaut et al 2003 and Lowe et al 2004 did report
reduction in length of stay for some women. It is possible that the inability to detect any
change in behaviour in some studies, was due in part, to the lack of a effective
educational intervention and the lack of an implemented fFN testing protocol – thus
physicians did not incorporate the fFN test results to change their management of preterm
labour.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
17
© 2006, Institute of Health Economics
6.0
Conclusions
The following summarize the conclusions of the review of the literature:
•
Based on its high negative predictive value, fFN testing has theoretical potential
to reduce health care utilization and unnecessary treatment by more accurately
identifying women who are not in true preterm labour.
•
The potential of fFN testing has been shown in non-experimental studies, but
given the study design, the change in behaviour may be due to factors that
influenced changing practice patterns, other than the introduction of the fFN test.
•
The potential to reduce health care utilization and unnecessary treatment has not
been confirmed, for the most part, in the experimental (RCT) studies that have
been conducted. These RCTs did not incorporate explicit protocols for positive
and negative fFN management, nor did they incorporate demonstrated educational
interventions. Knowledge translation is a key component of these RCTs however
this was not fully addressed by any of the trials. This may explain the limited
impact of fFN testing in terms of change in behaviour managing PTL.
•
Both the non-experimental and experimental evidence has flaws, which affect the
robustness of their conclusions.
•
The literature supports the importance of establishing a protocol for the use of
fFN results in the management of PTL, education both initially and on an ongoing
basis is required and audit is necessary to ensure the potential is realized. “..if
physicians are reluctant to abandon interventions in the setting of a negative test
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
18
© 2006, Institute of Health Economics
result and begin to use increased interventions after a positive test (which has a
low predictive value), the number of medical and social resources that are
consumed will actually increase after the use of the test.” Grobman et al, 2004.
•
There is a suggestion that using fFN testing is most useful in the context where
clinical criteria would warrant admission (Plaut et al, Sullivan et al, Abenhaim et
al). For those whom clinical criteria rules out PTL there is little benefit; there is
particular benefit for those at ‘high-risk”, i.e., those who would be admitted on
clinical criteria.
•
The published literature considers the impact of fFN testing available in referral
tertiary hospitals where admissions (or transfer) for care for PTL would occur.
The impact of fFN testing outside these settings has not been examined in the
published literature.
•
None of the studies considered the impact of fFN testing on quality of life, and in
particular maternal stress and anxiety. A study looking at using fFN testing to
predict preterm delivery in non-symptomatic high risk women (Shennan et al
2005) found that a positive fFN result was associated with statistically and
clinically significantly higher levels of anxiety. However, this is not likely
applicable to the use of the test in women experiencing signs and symptoms of
PTL who are already experiencing high levels of anxiety.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
19
© 2006, Institute of Health Economics
Part II: Analysis of cost implications of fFN testing for Alberta
7.0
Introduction
The objectives of Part II of this report are to describe the population characteristics of
Alberta women presenting with symptoms of PTL as well as to examine the cost
implications of fetal fibronectin testing in Alberta. The objectives of this study are
addressed by an analysis of administrative data from Alberta Health and Wellness.
8.0
Data extraction strategy from Alberta Health databases
We identified all individuals in the inpatient or ambulatory care databases with a
diagnosis of O42.xxx, O47.xxx, O75.xxx or O60.001 or grouping to CMG 599 or CMG
619. (See Appendix 3) We received demographic information and information on all
inpatient, ambulatory care and physician services received during the first encounter and
following that until fiscal year end.
(See Data requirements summary in Appendix 4).
We have received data year 2004/2005, and also data year 2002/20031. In 2004/2005, fFn
testing had already been implemented in the Calgary Health Region and in addition
during the period June 15, 2004 through Feb 15, 2005 a pilot study using the fFN test was
being conducted at the Royal Alex.
1
Fiscal year 2002/2003 has been received as of March 3, 2006, but is not included in this report. Note,
additional analyses for 2002/3 are not included in the primary analysis. However, where relevant this data
will be referred to.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
20
© 2006, Institute of Health Economics
The initial data extraction strategy involved a broad inclusion strategy.
Our initial
analysis will focus on three sub samples of women: those women with a diagnosis of
O47003 (threatened preterm labour <37 weeks gestation), or O75803 (preterm labour
with delivery delayed by therapy, antepartum condition) or O60000 (preterm birth). We
will further exclude those women whose gestational age at the time of diagnosis was >35
weeks, and who had ruptured membranes (042xxx) since the fFN test is not appropriately
used in these women.
9.0
Population Characteristics of women presenting with threatened preterm
labour
The following descriptive information is provided on the sub-sample of women who
present with symptoms of preterm labour and have an inpatient or outpatient encounter
with a coding of O47003. There were 1962 such women. We then exclude those from
whom we could not confirm when they gave birth (171 women), who had ruptured
membranes (144), who were at gestational week above 35 at first visit (297), who had
missing age of the mother and missing gestational age (5). We also excluded 99 women
who had an outpatient assessment for threatened preterm labour but did not have an
inpatient assessment – presumably these women were ruled out as being in true preterm
labour in the outpatient setting and thus fFN testing was not relevant. We ended up with
a sub-sample of 1247 women. This includes 212 women who were at gestational week
35 at their first visit. The project charter suggested that these women are part of the
target population although the Society of Obstetricians and Gynaecologists of Canada’s
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
21
© 2006, Institute of Health Economics
special report on Preterm labour refers to women less than 34 weeks gestation as the
target population.
We will also present similar information about a second sub-sample of women who had a
diagnosis of O60000 (PTB) but without having had a diagnosis coded O47000, as these
women are also women who presented to the system with signs of PTL. Because they
actually progressed to give birth, they may not have had a coding of threatened labour
although at their initial presentation they would have been considered in threatened
labour until the diagnosis was confirmed. These women will also be considered as
potential patients on whom the fFN test might be conducted. There were 2734 of these
patients. Those who had ruptured membranes were excluded (911), as were those who
had had a previous admission for threatened PTL and were captured in the previous subsample (294), and those who were at greater than gestational week 35 at the visit (651)
and those for whom there was no age/region of residence information (2). This subsample does include 247 patients at gestational week 35. This final sub-sample consists
of 844 women.
Finally, we consider the sub-group of women with a diagnosis of O75803 (preterm
labour, delivery delayed by therapy). There were 696 such women. After excluding
those for whom we could not confirm date of birth (76), those with ruptured membranes
(41), outpatient assessment without inpatient admission (5), no gestational age or
mother’s registry information (5), those already included in the other two sub samples
(273) those with gestational age greater than 35 at first visit (154). We also included in
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
22
© 2006, Institute of Health Economics
this group 149 patients from the O47.003 sub sample who had an outpatient admission
for O47.003, but never had an inpatient admission for O47.003. This resulted in a final
sub sample of 291, including 11 at gestational week 35.
9.1
Number of women
In 2004/2005 there were 1247 women diagnosed with threatened PTL in either an
outpatient or inpatient setting. In 2004, there were 40,581 births in the province so this
represents approximately 3% of births. There were also 846 preterm births in women who
never had an episode of threatened PTL, and another 280 with a diagnosis of preterm
labour delayed by therapy. In total, there were 2396 women who would have initially
presented to the system with symptoms of PTL. This represents 5.9% of births in 2004.2
In the 1247 women diagnosed with threatened PTL, 359 of them have an inpatient
admission for threatened PTL. There are 394 inpatient admissions in this group. Of the
888 patients with an outpatient admission for threatened PTL only, for all but one there is
an eventual birth related inpatient admission. 238 of these women had additional inpatient admissions prior to the birth admission – 139 of these were for 075.803 and as
noted these women we included in that sub-sample. The remainder of these admissions
were for other complications such as gestational hypertension, cervical incompetence,
hemmorage, or other diseases complicating pregnancy.
2
There were 40,581 births in Alberta in 2004-2005 (Source: Alberta Perinatal Health Program).
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
23
© 2006, Institute of Health Economics
In the 844 women with a PTB diagnosis, there are 846 inpatient admissions. There are
two women who have two preterm births during the one year period of our data.
In the 291 women with a diagnosis of preterm labour with delivery delayed by therapy,
there are 280 inpatient admissions. There are 11 women with an outpatient admission,
but no inpatient admission for this diagnosis.
9.2
Gestational age at birth
Women with threatened PTL presented for the first time between gestational week 12 and
35, with the average gestational age being 31 weeks. The average gestational age at birth
was 37 weeks. 73% of these women went on to give birth at term (>= 37 weeks), and
27% gave birth preterm.
22% of them were born between preterm between 33-36
weeks, 4% moderately preterm between 28-32 weeks and 1 % were extremely preterm
between 20-27 weeks.
Similarly, in the sub-sample of women with preterm labour
delayed by therapy, 80% delivered at term, 19% delivered between 33-36 weeks, and 1%
each were moderately or extremely preterm.
In the PTB sub-sample, the average gestational age was 32 weeks.
64% were preterm
between 33-36, 22% were moderately preterm between 28-32 weeks and 14% were
extremely preterm between 20-27 weeks.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
24
© 2006, Institute of Health Economics
Gestational age at Birth: PTL subsample
80%
70%
Percent
60%
50%
40%
30%
20%
10%
0%
At term (>=37)
Preterm (33-36)
Moderately preterm Extremely preterm
(28-32)
(20-27)
Gestational age at birth: PTL therapy subsample
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
At term (>=37)
Preterm (33-36)
Moderately
preterm (28-32)
Extremely preterm
(20-27)
Gestational age at birth: PTB subsample
70%
60%
50%
40%
30%
20%
10%
0%
At term (>=37)
Preterm (33-36)
Moderately
preterm (28-32)
Extremely
preterm (20-27)
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
25
© 2006, Institute of Health Economics
9.3
Regional distribution
For the 1247 women with threatened preterm labour, over 50% of them were from
Capital Health Region and only 11% were from the Calgary Health Region. However,
when you look at only the 394 women with inpatient admissions for threatened preterm
labour, we see 31% in Calgary and 28% in Edmonton. This is in comparison to Calgary
accounting for 36% of all live births and 38% of all preterm births, while Capital
accounts for 29% of all births and 31% of preterm births in 2003 (Alberta Reproductive
Health Report, 2005 Tables update). There are far more women in Capital who have an
outpatient assessment for threatened preterm labour, however roughly the same number
of women in the two regions who have admissions for threatened preterm labour.
For the group of women experiencing PTB without a previous PTL admission, Calgary
accounts for 44% and Capital accounts for 27%. We can also see that for the sub sample
with the diagnosis of PTL delayed by therapy, Calgary accounts for 20% and Capital
accounts for 37%.
Region of residence for women with threatened
PTL
Co
un
No
tr y
rth
er
n
Li
gh
ts
As
pe
n
Pe
ac
e
l
Ca
pi
ta
Pa
ll is
er
Ca
Da
lg
ar
vid
y
Th
om
ps
on
Ea
st
Ce
nt
ra
l
Ch
i
no
ok
60%
50%
40%
30%
20%
10%
0%
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
26
© 2006, Institute of Health Economics
Region of residence for women with threatened
PTL inpatient admissions
Pe
No
ac
e
rth
er
n
Li
gh
ts
As
pe
n
l
Ca
pi
ta
Pa
ll is
er
Ca
Da
lg
ar
vid
y
Th
om
ps
on
Ea
st
Ce
nt
ra
l
Ch
i
no
ok
35%
30%
25%
20%
15%
10%
5%
0%
Region of residence for women with PTL delayed
by therapy
Pe
ac
No
e
rth
er
n
l ig
ht
s
As
pe
n
l
Ca
pi
ta
ps
on
Ea
st
Ce
nt
ra
l
Th
om
ga
ry
Ca
l
Da
vid
Pa
ll is
er
Ch
i
no
ok
40%
35%
30%
25%
20%
15%
10%
5%
0%
Region of residence for women with PTB
50%
40%
30%
20%
Co
un
No
tr y
rth
er
n
Li
gh
ts
As
pe
n
Pe
ac
e
l
Ca
pi
ta
Pa
ll is
er
Ca
Da
lg
ar
vid
y
Th
om
ps
on
Ea
st
Ce
nt
ra
l
Ch
i
no
ok
10%
0%
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
27
© 2006, Institute of Health Economics
9.4
Age Structure
The average age of women presenting with PTL was 27, with the youngest being 15 and
the oldest being 44. The average maternal age for all births in Alberta in 2003 was 29.
Thus, this sub-sample of women presenting with PTL is a younger group. The group of
women with PTL delayed by therapy are similar with an average age of 27. In both
groups, there was no difference in the maternal age for those who went on to give birth at
term, verses those who did not. For the PTL sub-group, average maternal age by region
is shown below.
Average Maternal age by region of residency
(PTL sub-sample)
RHA
Age (years)
Chinook Regional Health Authority
24.6
Palliser Health Region
28.3
Calgary Health Region
27.7
David Thompson Regional Health Authority
25.8
East Central Health
27.3
Capital Health
28.0
Aspen Regional Health Authority
25.1
Peace Country Health
25.4
Northern Lights Health Region
26.5
The average age of women in the PTB sub-sample is 30 years, with the youngest being
14 and the oldest being 46. These women are slightly older than the average maternal
age for all births in Alberta.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
28
© 2006, Institute of Health Economics
9.5
Socioeconomic Structure
The indicator of socioeconomic status available in the data is whether the mother is
aboriginal, receives health care premium support or is on welfare. In the sub sample of
women presenting with PTL, 10% are aboriginal, 16% receive premium support and 9%
are on welfare.
The sub sample with PTL delayed by therapy is very similar. The
aboriginal and people income support groups are overrepresented in this sub sample of
women compared with the female population between 13 and 50, or with distribution of
children under the age of 1. In particular, those on welfare which is 9% of the women
with threatened PTL, but account for only 3% of childbearing women and 3.5% of
children less than one.
Also 10% of the women with threatened PTL are aboriginal
compared to 4% and 6% of the childbearing women and children under 1 respectively.
The group of women with PTB more closely mimic the population of childbearing
women or of children under age 1, but are still over represented in the aboriginal,
premium support of welfare groups. This disparity is more evident though in the group
of women with threatened preterm labour.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
29
© 2006, Institute of Health Economics
Women Presenting with Preterm Labour 2004/2005
SES
Count
0
A
S
W
Total
823
131
195
108
1257
% Distribution
65.5%
10.4%
15.5%
8.6%
100.0%
645
72
83
46
846
% Distribution
76.2%
8.5%
9.8%
5.4%
100.0%
Women with Preterm Birth 2004/2005
SES
Count
0
A
S
W
Total
Female Population, Age >= 13 and <=50, Alberta, March 31, 2005
SES
0
A
S
W
Total
Active Population Count (in 000's)
751.8
37.1
106.5
27.0
922.4
% Distribution
81.5%
4.0%
11.5%
2.9%
100.0%
Source: Population Registry File 2004/2005
Female and Male Population, Age < 1, Alberta, March 31, 2005
SES
0
A
S
W
Total
Active Population Count (in 000's)
31.3
2.5
5.3
1.4
40.5
% Distribution
77.3%
6.2%
13.1%
3.5%
100.0%
Source: Population Registry File 2004/2005
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
30
© 2006, Institute of Health Economics
10.0
Cost implications
10.1
Costs of fFN testing
As per the AHFMR fFN report, based on communication with the manufacturer3:
Specimen Collection Devices, consisting of a plastic tube with buffer, cap and swab are at
no charge. Dating on the product is 4 years at room temperature. It is possible to break
the specimen collection kits into individual units for small testing sites.
Test cartridges for fFN testing are in boxes of 24 and are $97.27 per determination and
have an 18-month dating from manufacture. It is possible to break the specimen
collection kits into individual units for small testing sites.
The Tli IQ instrument system and printer which consists of all hardware components to
run an fFN assay on site. For hospitals >1000 births/year, the hardware components are
supplied at no charge. For hospitals <1000 births /year, the equipment is available on a
rent-to-own basis for a 24 month payment of $111 for 24 months or purchased for $2668.
Currently, all service, replacement and training costs are included.
These prices include all quality control of both the test cartridges and the Tli IQ.
3
This updates the costing information available in the Alberta Health Technologies Decision Process,
Fetal Fibronectin Project Charter. The per determination cost is US$100, and the instrument system is
US$2400. This has been converted to Canadian dollars at a rate of $1.11.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
31
© 2006, Institute of Health Economics
10.2
Potential number of tests
Using administrative data on the number of admissions related to preterm labour
symptoms, we can estimate the potential number of fFN tests that might be conducted if
it were available. We have considered two different populations of patients as potentially
eligible for the fFN test, and also considered the possibility that is available in different
types of hospitals.
Potential number of tests done if all admissions for women in the various sub-sample
groups are included:
•
394 inpatient admissions for women with threatened PTL diagnosis
•
846 inpatient admissions for women with PTB diagnosis
o Note – this is likely an overestimate as for some of these patients it
may be obvious early that the symptoms of labour are progressing
and an fFN test would not be done.
o 280 inpatient admissions for women with PTL delayed by therapy
¾ Potential number of tests: 1520
Another possibility is that fFN testing is only done at Level II or Level III hospitals
where treatment for possible preterm birth would be provided.
The Level III hospitals
are Foothills hospital in CHR and Royal Alexandra in Capital. The Level II hospitals are
Lethbridge Regional in Chinook, Rockyview and Peter Lougheed in Calgary, Red Deer
General in David Thompson, Misericordia and Grey Nuns in Capital, Queen Elizabeth II
in Peace.
For the purposes of this analysis, we have also included Medicine Hat
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
32
© 2006, Institute of Health Economics
Regional Health Centre in Palliser. In this case, the number of potential tests is as
follows:
•
321admissions for threatened PTL
•
819 for PTB
•
265 for PTL delayed by therapy
¾ Potential number of tests: 1405
10.3
Variable costs of fFN testing
This section considers the variable cost of the fFN test cartridges. The laboratory costs
for processing are not included. There are two different scenarios considered in terms of
the number of fFN tests that might be performed (see table below). The first (Scenario 1)
is to assume that one test per inpatient admission would be performed and consider each
of our three sub samples (those with PTL, those with PTL delayed by therapy and those
with PTB) as the population on whom the test would be done. The variable cost of
testing at $100 per test, then would range between $38,324 and $147,850.
The other clinically relevant scenario is that the fFN test is only available at Levels II and
III hospitals. If a woman it suspected of being in preterm labour then she would be
transferred and admitted to one of these facilities depending on her the gestational week
of her pregnancy. If in both Level II and III hospitals, the cost of testing ranges between
$31,224 and $136,664.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
33
© 2006, Institute of Health Economics
Scenarios
fFN testing in all facilities
Variable cost of fFN testing
Number
of tests
PTL only (394)
above + PTL delay by tx (280)
above + PTB (846)
Variable costs of testing
($97.27 per test)
394
674
1520
$38,324
$65,560
$147,850
321
586
1405
$31,224
$57,000
$136,664
fFN test at Level II and III hospitals
PTL only (321)
above + PTL delay by tx(586)
above + PTB (819)
10.4
Fixed costs
If only implemented in Levels II and III hospitals, there would be no fixed costs because
the fFN hardware costs for hospitals with greater than 1000 births is provided at no
charge.
If implemented in all hospitals: there are 62 Level I hospitals where births occurred in
2004. The total hardware costs, at $$2668 per, would thus be $162,748.
10.5
Length of Stay
The average length of stay (LOS) for threatened PTL was 2.01 days, ranging between 1
and 69. Excluding two outliers (LOS > 30 days), the average LOS for admissions related
to threatened PTL was 1.77.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
34
© 2006, Institute of Health Economics
The average length of stay for PTL delayed by therapy was 3.44 days, ranging between 1
and 48. Excluding outliers, the average LOS was 2.87.
The average length of stay related to preterm birth is 5.33 days, ranging between 1 and 50
days.
Recall that fFN testing had been implemented in the Calgary Health Region during the
time period covered by this data. In the Calgary Health Region, the average length of stay
for threatened PTL is 1.82 days. In the Capital Health Region, the average length of stay
is 2.69. Not including the Royal Alex, the average length of stay in Capital is 3.5. The
average length of stay for threatened preterm labour is shorter in the Calgary Health
Region during this period (when fFN testing is being done) than in Capital. This should
be interpreted with some caution, as there may be differences in the case-mix or other
influencing factors between these two Health regions not related to fFN testing.
Comparing LOS in the two regions in the 2003/2004 data shows that the LOS in Capital
was also higher than Calgary without fFN testing, and that in both cases LOS fell over
time. The rate of decline was larger in Calgary which may be associated with the
availability of fFN testing, although this association can not be confirmed with this data.
Also, during the data year 2004/2005, Capital Health region was undertaking a pilot
study of fFN testing at the Royal Alexandra. The fFN pilot study was going on during
June 15, 2004 through February 15, 2005.
The average length of stay during the fFN
pilot study at Royal Alex was 1.85 days (excluding two outliers who had LOS of 30 and
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
35
© 2006, Institute of Health Economics
69 days for which there were complicating factors determining the LOS). During the
non-fFN availability at the Royal Alex, the average LOS was 2.91. Thus, the average
LOS at the same hospital was a day shorter when the fFN test was available.
10.6
Average Costs for threatened PTL
9.6.1
Inpatient admission
The provincial average cost was $669 per inpatient admission day.
9.6.2 Physician billing
The average physician billing was $108 per inpatient admission day.
9.6.3 Potential costs avoided from LOS reductions
Consider the scenario of putting fFN testing into Level III and II hospitals only. There are
1405 admissions, including 819 for PTB. Using the information on average LOS at the
Royal Alexandra as an indicator, we will assume that fFN testing reduces overall length
of stay by 1 day. This is a conservative estimate – for example, the Calgary study showed
a greater reduction in length of stay for specific admissions (see the AHFMR fFN report)
Excluding those admissions that include PTB, this would result in a potential to decrease
LOS then for 586 admission days. The hospitals in the CHR already have fFN testing –
thus based on the number of admissions we assume that each would have lasted an
additional day, and thus that there would have been an additional 194 admission days.
These are admission days avoided which have an opportunity cost of $150,554. In
Capital Health, there is a potential for a reduction of 301 admission days. This represents
a savings valued at $233,576. In the other Level II hospitals outside of Capital and
Calgary, there are a potential for a reduction of LOS for 91 admission days. However, the
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
36
© 2006, Institute of Health Economics
average LOS for these admissions is already very close to 1, so we conservatively
estimate that only half will have a 1 day reduction. This results in LOS reduction valued
at $35,308. Thus, the total potential costs avoided due to LOS reductions is $419,428.
If instead we consider the scenario where fFN testing is available in all hospitals, the
same LOS reductions as noted above would still be applicable. In addition, there would
be any potential LOS reductions from the availability of the test in Level I hospitals.
There are 115 admission days, including 27 for PTB in Level I hospitals. Thus, there are
88 admissions that could potentially have an LOS reduction. If we assume a 1 day
reduction, the value of the LOS reduction is $68,288. However, again the average LOS
for these is very close to one so a conservative estimate with only half of the admissions
having 1 day reduction would results in a cost savings for LOS reduction of $34,144.
This results in total potential costs avoided due to LOS reductions in this scenario of
$419,428 if no additional LOS reductions in the Level I hospitals are possible or between
$453,572 to $487,716 if some reductions are possible.
10.7
Ambulance Transfers
10.7.1 Numbers of Transfers between facilities
There were 64 ambulance transfers in this group of women with threatened PTL
associated with admissions for threatened PTL. Of these, only 1 resulted in birth. There
were an additional 42 transfers between facilities that did not involve the use of an
ambulance. Presumably, the patients arranged their own transportation in these cases. Of
these 5 resulted in birth.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
37
© 2006, Institute of Health Economics
Looking at the 64 ambulance transfers only, there were 53 that were by ground
ambulance, 4 by air ambulance, and 7 by combined air and ground ambulance.
All four of the air ambulance transfers are to the Royal Alexandra in Capital Health
Region. Three of these are from Aspen Regional Health Authority, and one is from
Peace.
Looking at the combination transfers, 6 of them involve transfers between regions and 1
represents a transfer within the Northern Lights health region. The between region
transfers are as follows: There was one transfer from Palliser to Calgary Health Region,
two from Aspen to Capital, one each from Northern Lights and Peace to Capital, and one
from Capital back to Peace. The first case was a transfer to Foothills from Palliser for an
admission that resulted in birth – therefore is an unavoidable transfer. The latter case is a
woman who was transferred to Royal Alex from Peace River via ground ambulance (but
not coded for PTL) and then transferred back to Peace River by combined air/ground
ambulance and discharged. She gave birth a month later at 40 weeks gestational age, in
Peace River. These two cases are removed from consideration.
For the ground ambulance transfers, 17 of the 53 were between health regions. Some of
these ground ambulance transfers would be unavoidable because they came from the
community. Removing the ‘unavoidable’ ground ambulance transfers, there are 38
ground ambulance trips.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
38
© 2006, Institute of Health Economics
Ambulance Transfers (for threatened PTL subgroup)
Inter-regional Transfers
From RHA
To RHA
Air
Ground
Combo
Comments
Palliser
Calgary
1
1*
unavoidable
David Thompson
Calgary
2
Northern Lights
East Central
Aspen
Peace
David Thompson
Capital
Capital
Capital
Capital
Capital
Capital
Total inter-regional
Peace
3
1
4
1
4
7
1
1
17
1
2
1
1*
6
unavoidable
Intra-regonal Transfers
Air
Calgary
Capital
Northern Lights
Aspen
Chinook
Palliser
Peace
David Thompson
Total intra-regional
Total trips
Total avoidable
10.7.2
Combo
0
Ground
8
15
2
2
3
2
1
3
36
4
4
53
38
7
4
1
4 unavoidable
2 unavoidable
all unavoidable
2 unavoidable
all unavoidable
one unavoidable
one unavoidable
one unavoidable
1
64
48
Total ambulance Transfers
We also looked at all air or combined ambulance transfers in the entire data set not just
for the PTL sub-sample. There were 200 of these. For 57 of them, the gestational week
was >35 thus fFN testing is not applicable. There were 55 trips with a gestational age <35
and these were examined – they were all for preterm births, thus would not be avoidable
transfers. For the remaining 88 of them, the gestational age was missing requiring
additional investigation. Looking more closely at the missing GA trips, it turns out there
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
39
© 2006, Institute of Health Economics
are 8 avoidable air transfers, and 9-13 avoidable combination air/ground transfers. (See
Appendix 5 for details).
10.7.3 Costs of ambulance transfers
Average air ambulance cost: $3644
Average ground ambulance cost: $450-500 (from IHE Provincial Cost List)
10.7.4 Cost savings from avoided ambulance transfers
Using the information from the analysis of PTL transfers only:4
•
8 avoidable air/combined ambulance trips - $29,152 savings
•
42 ground ambulance/combined trips – $21,000 savings
Using the more complete information from the detailed examination of at all
air/combination transfers, there were 8 potentially avoidable air transfers and 9-13
potentially avoidable air/ground combination transfers:
•
11.0
$77,491 - $87,781 in potentially avoidable costs.
Overall cost implications
Putting all the information on the cost implications from the previous sections together,
we consider the costs of fFN testing, including both variable and fixed, and then offset
these with the costs avoided from the value of LOS reductions and avoided ambulance
transfers to consider the overall cost implications. Note that the bulk of the costs avoided
are in reduced LOS – thus do not represent dollars released, but do address capacity
issues. See Appendix 6 for detailed calculations.
4
This does not include the PTL delayed by therapy sub-group, and this is conservative.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
40
© 2006, Institute of Health Economics
For the base case (scenario 2) with testing at Levels II and III hospitals:
¾ Net costs avoided are $280,596
o The bulk of this savings are realized in Capital and Calgary
¾ Assumes no air ambulance transfers prevented
¾ This includes one ground ambulance transfer prevented
¾ This assumes doing the fFN test on all admissions which results in
PTB – this is likely an overestimate and thus the net savings are
conservative.
Consider also scenario 1 (test in all facilities):
¾ Overall cost implications range from an overall cost avoidance of
$207,330 to $307,118.
o Adding testing to Level I facilities as well means overall savings
are reduced in all but the least conservative sub-scenario compared
with Scenario 2 (see Appendix 6).
¾ This assumes all air/combined ambulance transfers avoided are allocated
to this Scenario.
¾ The sub-scenarios vary in terms of the assumed amount of the ambulance
transfers. This assumes ground ambulance transfers range between
$21,000 based on using the information from the PTL transfers only plus
an assumed $42,000 to account for if all PTL and PTL delayed by therapy
were included (ground ambulance transfers for all patients have not been
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
41
© 2006, Institute of Health Economics
examined in detail, only the PTL subgroup). Air transfers avoided range
between $77,500 and 88,000).
¾ This assumes doing the fFN test on all admissions which results in PTB –
this is likely an overestimate and thus the net savings are conservative.
¾ The sub-scenarios vary in terms of the potential for admission day
reductions outside the Level II and III hospitals. There are an additional
115 admissions in 36 Level I facilities in this fiscal year – including 27
admissions for PTB. The average number of admissions per facility is
2.75, with a range between 1 and 14. Given that the average LOS is
already less than 2 days, it could reasonably be assumed these are not
reduced any further. (Scenario 1a, 1b) Further, given that these patients are
not being transferred to higher level facilities one could assume that they
are being correctly diagnosed in their home facilities. Alternately, you
could assume there is a 1 day reduction, or more conservatively that 50%
of admissions have a 1 day reduction.
These detailed cost calculations consider the impact of fFN testing in a one year horizon
only, and assume that the fixed costs of purchasing the equipment are all borne in year 1.
However, it is important to note as well that in Year 2, the net savings would then be
increased in Scenario 1 by the amount of the fixed costs: thus adding $162,748 to each of
the net cost savings in Scenarios 1a-1f. Assuming no change in the number of tests and
admission days, this would mean that fFN testing in the Level I facilities would result in
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
42
© 2006, Institute of Health Economics
additional system savings in year 2 for all the sub-scenarios (ranging between $89,482
and $189,270).
Alternately, looking at a two year horizon and assuming that the equipment is leased on a
48 month basis the fixed costs would be equally split between the two years, and in all
but the sub-scenarios, there are additional system savings in both years of between
$8,108 and $107,896.
12.0
Conclusions
There are a few limitations and caveats to note regarding this analysis based upon
administrative data. First, the number of tests and reductions in admissions is estimated
using administrative data, thus the actual numbers of both are not known. A conservative
approach was taken both in terms of the number of tests (we included admissions for
PTB in estimating potential number of tests) and also in estimating possible length of stay
reductions. It is important to note that only Alberta residents only are included in the
data analysis, thus out of province admissions to the tertiary care facilities are not
accounted for. In this analysis, the impact of fFN testing only considers reduced length
of stay, and there is no way to account for the important impact of reductions in
unnecessary treatments (not just costs of these treatments, but the benefits associated with
women not receiving these drugs unnecessarily). Out of pocket costs of families are not
included, and in addition the anxiety and stresses associated with the experience of
preterm labour, and how this may be impacted by fFN testing, are not accounted for. The
impact of fFN testing in Level I facilities is speculative as there is little information
available in the literature to base projected impact upon.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
43
© 2006, Institute of Health Economics
Based on the analysis of the administrative data, the following conclusions regarding the
population of women in Alberta with signs of PTL:
•
The number of women presenting with signs of PTL represent approximately 6%
of all births in Alberta.
•
The majority of these do not progress to PTB.
•
Threatened PTL disproportionately affects more disadvantaged groups.
The following conclusions regarding the impact of fFN can be drawn:
•
There are shorter average LOS for threatened PTL when fFN is available than
when it is not.
•
There were 43 ground ambulance transfers between hospitals that could have
been avoided – of these the majority are within regions, and only 19 involve
between region transfers.
•
An examination of air/combined transfers reveals a small number of potentially
avoidable transfers: 8 air transfer and 9-13 combination transfers (out of a total of
200).
•
Implementing fFN testing in Level II and III hospitals involves net cost “savings”
to the system, primarily driven by cost avoidance from reduced LOS.
•
Implementing fFN testing in Level I hospitals involves additional system savings,
when a two year time horizon is considered. However, there is little evidence in
the literature on the use of fFN testing in these types of facilities, therefore these
conclusions should be further researched.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
44
© 2006, Institute of Health Economics
References
Abenhaim HA, Morin L, Benjamin A. Does availability of fetal fibronectin testing in the
management of threatened preterm labour affect the utilization of hospital resources?
Journal of Obstetrics & Gynaecology Canada: JOGC 2005; 27(7):689-694.
Corabian P and Harstall C. The role of the rapid fetal fibronectin assay in the management of
suspected preterm labour. Alberta Heritage Foundation for Medical Research, 2006.
Joffe GM, Jacques D, Bemis-Heys R, Burton R, Skram B, Shelburne P. Impact of the fetal
fibronectin assay on admissions for preterm labor. American Journal of Obstetrics and
Gynecology 1999; 180(3 PART 1):581-586.
Giles W, Bisits A, Knox M, Madsen G, Smith R. The effect of fetal fibronectin testing on
admissions to a tertiary maternal-fetal medicine unit and cost savings. American Journal of
Obstetrics & Gynecology 2000; 182(2):439-442.
Grobman WA, Welshman EE, Calhoun EA, Ramsey PS. Fetal fibronectin results did not reduce
medical resource use for women with preterm uterine contractions. American Journal of
Obstetrics and Gynecology 2004; 191, 235-40. .
Lowe MP, Zimmerman B, Hansen W. Prospective randomized controlled trail of fetal fibronectin
on preterm labour management in a tertiary care centre. American Journal of Obstetrics
and Gynecology 2004; 190(2): 358-62.
Musaad SMA, Melson CL, Boswell DR. Assessment of the impact of introducing fetal
fibronectin assay in the management of preterm labour at Middlemore Hospital, New
Zealand. Pathology 2005; 37(3):226-230.
Mozurkewich E, Naglie G, Krahn M, Hayashi R. Predicting preterm birth: a cost-effectiveness
analysis. Am J Obstet Gynecol 2000;182(6): 1589-1598.
Norman JE and Greer IA, Preterm Labour: managing risk in clinical practice. New York:
Cambridge University Press, 2005.
Nguyen TCQ. The cost-effectiveness of fetal fibronectin testing in suspected preterm labor: a
randomized trial [abstract]. Obstetrics & Gynecology 2002; 99(4 Suppl):97S, 2002.
Plaut MM, Smith W, Kennedy K. Fetal fibronectin: the impact of a rapid test on the treatment of
women with preterm labour symptoms. American Journal of Obstetrics and Gynecology
2003; 188(6): 1588-1593.
Shennan A, Jones G, Hawken J, Crawshaw S, Judah J, Senior V, Marteau T, Chinn S, Poston L.
Fetal fibronectin test predicts delivery before 30 weeks of gestation in high risk women, but
increases anxiety. BJOG: an International Journal of Obstetrics and Gynacecology,
2005;112: 293-298.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
45
© 2006, Institute of Health Economics
Sullivan A, Hueppchen NA, Satin AJ. Cost effectiveness of bedside fetal fibronectin testing
varies according to treatment algorithm. Journal of Maternal-Fetal Medicine 2001;
10(6):380-384.
Reproductive Health Report Working Group. Alberta Reproductive Health: Pregnancies and
Births 2005 Tables Update. Edmonton AB: Alberta Health and Wellness.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
46
© 2006, Institute of Health Economics
APPENDIX 1
SEARCH STRATEGIES
Premature Labour Economics Literature Search
MEDLINE (OVID 1966 to October Week 2 2005)
Cochrane Central Register of Controlled Trials (OVID 4th Quarter 2005)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
fibronectins or receptors, fibronectin[MeSH exploded terms]
labour, premature[MeSH exploded]
(preterm or premature) ADJ2 (labour or labour or birth or deliver*)[Text Words]
2 or 3
1 and 4
(fetal or foetal) ADJ2 fibronectin*[Text Words]
5 or 6
labour, premature/economics[MeSH exploded]
costs and cost analysis or economics[MeSH exploded terms]
cost or costs or cost effective* or economic*[Text Words]
9 or 10
4 or 7
11 and 12
8 or 13
limit 14 to English language
labour, premature/diagnosis[MeSH Major exploded]
limit 1 to review articles
Canada[MeSH exploded]
Canada or Canadian or British Columbia* or Alberta* or Saskatchewan* or Manitoba* or
Ontario* or Quebec* or Nova Scotia or New Brunswick or Newfoundland or Prince Edward
island or PEI[Text Words]
3 or 4
1 and 5
trend* or history or historical or utilization or utilisation[Text Words]
1 and 7
labour, premature/history[MeSH exploded]
2 or 6 or 8 or 9
limit 10 to human and English language
EMBASE (OVID 1980 to 2005 Week 42)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
fibronectin or fibronectin binding protein or fibronectin receptor[EMBASE exploded terms]
premature labour[EMBASE exploded term]
(preterm or premature) ADJ2 (labour or labour or birth or deliver*)[Text Words]
2 or 3
1 and 4
(fetal or foetal) ADJ2 fibronectin*[Text Words]
5 or 6
economic evaluation or economic aspect[EMBASE exploded terms]
cost or costs or cost effective* or economic*[Text Words]
8 or 9
4 or 7
10 and 11
limit 12 to English language
premature labour/diagnosis[EMBASE Major exploded term]
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
47
© 2006, Institute of Health Economics
15. limit 1 to reviews
16. Canada[EMBASE exploded]
17. Canada or Canadian or British Columbia* or Alberta* or Saskatchewan* or Manitoba* or
Ontario* or Quebec* or Nova Scotia or New Brunswick or Newfoundland or Prince Edward
island or PEI[Text Words]
18. 3 or 4
19. 1 and 5
20. trend* or history or historical or utilization or utilisation[Text Words]
21. 1 and 7
22. 2 or 6 or 8
23. limit 9 to human and English language
CINAHL (OVID 1982 to October Week 2 2005)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
fibronectins[CINAHL exploded term]
labour, premature[CINAHL exploded term]
(preterm or premature) ADJ2 (labour or labour or birth or deliver*)[Text Words]
2 or 3
1 and 4
(fetal or foetal) ADJ2 fibronectin*[Text Words]
5 or 6
labour, premature/economics[CINAHL exploded]
economics or costs and cost analysis[CINAHL exploded terms]
cost or costs or cost effective* or economic*[Text Words]
9 or 10
4 or 7
11 and 12
13 or 8
limit 14 to English language
labour, premature/diagnosis[CINAHL Major exploded term]
limit 2 to review articles
Canada[CINAHL exploded term]
Canada or Canadian or British Columbia* or Alberta* or Saskatchewan* or Manitoba* or
Ontario* or Quebec* or Nova Scotia or New Brunswick or Newfoundland or Prince Edward
island or PEI[Text Words]
3 or 4
1 and 5
trend* or history or historical or utilization or utilisation[Text Words]
1 and 7
labour, premature/history, trends[CINAHL exploded terms]
2 or 6 or 8 or 9
limit 10 to human and English language
Cochrane Database of Systematic Reviews (OVID 3rd Quarter 2005)
1.
2.
3.
4.
5.
6.
7.
(preterm or premature) ADJ2 (labour or labour or birth or deliver*)[Keywords/Text Words]
(fetal or foetal) ADJ2 fibronectin*[Keywords/Text Words]
cost or costs or cost effective* or economic*[Keywords/Text Words]
1 or 2
3 and 4
(preterm or premature) ADJ2 (labour or labour or birth or deliver*)[Keywords/Text Words]
diagnose$ or test$ or predict$ or detect$ or screen$ or marker$[Keywords/Text Words]
NHS Economic Evaluation Database (University of York CRD)
Health Technology Assessment Database (University of York CRD)
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
48
© 2006, Institute of Health Economics
DARE Database of Abstracts of Reviews of Effects (University of York CRD)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
fibronectins or receptors, fibronectin[Subject terms]
labour, premature[Subject term]
(preterm or premature) AND (labour or labour or birth or deliver*)[Text Words]
2 or 3
1 and 4
(fetal or foetal) AND fibronectin*[Text Words]
costs and cost analysis or economics[Subject terms]
cost or costs or cost effective* or economic*[Text Words]
7 or 8
4 or 5
9 and 10
6 or 11 labour, premature[Subject term]
diagnose$ or test$ or predict$ or detect$ or screen$ or marker$[Text Words]
1 and 2
EconLit (EBSCO 1969 to present)
Biological Abstracts (ERL WebSPIRS 1980 to August 2005)
ProQuest Dissertations and Theses
Canadian Research Index (Microlog)
KUUC Knowledge Utilization Database (University of Laval)
1.
2.
3.
4.
5.
6.
7.
(preterm or premature) AND (labour or labour or birth or deliver*)[Keywords/Text Words]
(fetal or foetal) ADJ2 fibronectin*[Keywords/Text Words]
cost or costs or cost effective* or economic*[Keywords/Text Words]
1 or 2
3 and 4(preterm or premature) AND (labour or labour or birth or deliver*)[Keywords/Text Words]
diagnose$ or test$ or predict$ or detect$ or screen$ or marker$[Text Words]
1 and 2
Grey literature web sites search terms (see attached list of web sites)
1.
2.
3.
4.
5.
6.
7.
8.
(preterm or premature) AND (labour or labour or birth or deliver*)[Text Words]
(fetal or foetal) ADJ2 fibronectin*[Text Words]
cost or costs or cost effective* or economic*[Text Words]
1 or 2
3 and 4(preterm or premature) AND (labour or labour or birth or deliver*)[Text Words]
diagnose$ or test$ or predict$ or detect$ or screen$ or marker$[Text Words]
1 and 2
Canada or Canadian or British Columbia* or Alberta* or Saskatchewan* or Manitoba* or
Ontario* or Quebec* or Nova Scotia or New Brunswick or Newfoundland or Prince Edward
island or PEI[Text Words]
9. 3 and 4
10. trend* or history or historical or utilization or utilisation[Text Words]
11. 3 and 6
12. 5 or 7
NOTES:
a)
b)
c)
d)
denotes truncation
ADJ denotes adjacency
denotes truncation
ADJ denotes adjacency
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
49
© 2006, Institute of Health Economics
Preterm Labour Literature Search Web Sites
Note: Government web sites not searched as these will be covered elsewhere
Canadian Web Sites
Alberta Heritage Foundation for Medical Research
www.ahfmr.ab.ca
Canadian Coordinating Office for Health Technology Assessment
www.ccohta.ca/default.cfm
Canadian Health Services Research Foundation
www.chsrf.ca
Canadian Institutes of Health Research
www.cihr-irsc.gc.ca
Canadian Nurses Association
www.cna-nurses.ca/cna
Centre for Health Services and Policy Research, UBC
www.chspr.ubc.ca
Manitoba Centre for Health Policy
www.umanitoba.ca/centres/mchp/lmchp.htm
Canadian Healthcare Association
http://www.cha.ca/
Canadian Institute of Health Research:
http://secure.cihi.ca/cihiweb/
Canadian Medical Association
http://www.cma.ca
Canadian College of Health Services Executives
http://www.cchse.org/
Health Quality Council (Saskatchewan)
http://www.hqc.sk.ca
Institute for Clinical Evaluative Science
http://www.ices.on.ca/
Institute of Health Economics
http://www.ihe.ca
McGill University Health Centre. Health Technology Unit
http://www.mcgill.ca/tau
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
50
© 2006, Institute of Health Economics
Australia/New Zealand
New Zealand Health Information Service
http://www.nzhis.govt.nz
Centre for Health Economics (Monash University)
http://www.buseco.monash.edu.au/centres/che/
Centre for Health Program Evaluation (Monash University)
http://www.chpe.buse.co.monash.edu.au
Australian Policy Online
www.apo.org.au
Centre for Health Economics Research and Evaluation
www.chere.uts.edu.au
Monash Institute of Health Services Research
http://203.94.147.62/index.asp
UK Web sites
Centre for Health Economics (University of York)
http://www.york.ac.uk/inst/che
Centre for Reviews and Dissemination, University of York (UK)
http://www.york.ac.uk/inst/crd/welcome.htm
Kings Fund
http://www.kingsfund.org.uk
UK National Research Register
http://www.nrr.nhs.uk/
National Coordinating Centre for Health Technology Assessment
www.ncchta.org/index.htm
UK National Institute for Clinical Excellence
www.nice.org.uk
US Web sites
Health Policy Institute (Georgetown University)
http://www.georgetown.edu/research/ihcrp/index.html
RAND Organization
http://www.rand.org
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
51
© 2006, Institute of Health Economics
U.S. Agency for Healthcare Research and Quality
http://www.ahcpr.gov
U.S. Office of the Surgeon General (US)
http://www.surgeongeneral.gov
Milbank Memorial Fund
www.milbank.org
National Bureau of Economic Research
www.nber.org
National Institutes of Health
www.nih.gov
Social Science Research Network
www.ssrn.com/index.html
U.S. General Accounting Office
www.gao.gov
International Websites
World Health Organization
www.who.int/en/
International Health Policy Library
www.policylibrary.com/health
European Centre for Health Policy
www.who.dk/echp
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
52
© 2006, Institute of Health Economics
APPENDIX 2
SUMMARY TABLES FOR ECONOMICS ARTICLES
Abenhaim HA, Morin L, Benjamin A. Does availability of fetal fibronectin testing in the management of threatened preterm labour affect
the utilization of hospital resources? Journal of Obstetrics & Gynaecology Canada: JOGC 2005; 27(7):689-694.
Overall conclusion
Availability of fFN associated with a reduction in hospital admissions, length of stay and overall hospital costs in the
management of threatened PTL.
Aim
“To examine how the availability of fetal fibronectin testing affects the utilization of hospital resources”
Setting
Canada, Royal Victoria Hospital, a university-associated tertiary care hospital in Montreal.
Design
Prospective cohort of women presenting with possible PTL when fFN testing was available, compared with a
historical cohort of women prior to availability of fFN.
Methods
Inclusion: singleton pregnancies, presenting between 24-34 weeks gestation with signs and symptoms of PTL.
Comparison of rates, duration and costs of hospitalization for the two cohorts of women, with and without available
fFN testing. Comparison was made during a 20 week study period. For fFN group, this was six months after test
became available.
Subjects for both cohorts identified by review of birthing centre triage visits, admissions and discharges. fFN test
results taken from fFN testing log book.
Initial clinical evaluation same in both cohorts.
fFN protocol: Testing was not performed in women with multiple pregnancy, ruptured membranes,
vaginal bleeding, history of recent intercourse, recent digital examination. Specimens obtained as per
manufacturer instructions, discarded if PTL were clinically confirmed (cervix dilated > 3cm) or ruled out
(cervix closed and uneffaced and contractions ceased). fFN results available within 30 minutes (implying
point of care availability).
Management decisions were based on initial clinical evaluation and fFN test when indicated. Final decisions
regarding discharge for fFN negative made by attending physician.
Results
Patients were either (1) admitted and discharged undelivered (PTL), (2) delivered after admission (PTB), or (3)
discharged.
Characteristics of two cohorts were comparable in terms of number of subjects, and distribution of gestational age at
presentation.
12 inappropriate tests were done: 6 for twin pregnancies, two for inappropriate gestational age, and four for
‘reassurance’ of mother presenting with diagnoses other than PTL.
Authors point out that more women received “active management” (testing or admission) in study period (29+24 =
53) than baseline period (0+37). They characterize this as more “unnecessary” testing.
Critical comments
Outcomes: lower admission rate for PTL with eventual discharge undelivered in fFN group. Significantly lower
days hospitalized (5.2 vs. 0.6), lower cost of admission, lower total cost.
Of note is the fact that 25% of fFN tests were “inappropriate”. The authors did include these costs in their analysis.
However, they note the need for continued monitoring and education around appropriate use.
They excluded costs of subjects that delivered during admission – but these numbers weren’t significantly different
in the two groups.
Good that they waited to examine impact after a settling in period.
It is important to note that not ALL patients in fFN available period were tested (only 35 of 116), the rest were
admitted or discharged based on clinical criteria. If ALL had the test, it may not have been cheaper.
They authors point out…”the importance of reviewing and evaluating the use of fFN testing in the labour and
delivery ward to maintain quality assurance and to ensure that clinical practices are current and evidence-based.”
No neonatal outcomes considered.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
53
© 2006, Institute of Health Economics
Joffe GM, Jacques D, Bemis-Heys R, Burton R, Skram B, Shelburne P. Impact of the fetal fibronectin assay on admissions for preterm
labor. American Journal of Obstetrics and Gynecology 1999; 180(3 PART 1):581-586.
Overall conclusion
fFN use resulted in reduced PTL admissions, LOS, and prescriptions for tocolytics with no negative impact on
neonatal outcomes.
Aim
“..to determine whether the introduction of routine fetal fibronectin bedside testing affected costs and transfer rates
from referral district hospitals to a tertiary obstetric hospital, as well as direct admissions to a tertiary referral
hospital”
Setting
Australia. Tertiary obstetric hospital as well as 9 referring district hospitals.
Some fFN tests were done in physician’s offices.
Design
Prospective audit of fFN use in above settings, for 18 months (done alongside an RCT for a tocoyltic). Compared to
a baseline year pre-fFN.
Inclusion: threatened preterm labour. 24-34 weeks gestational age, cervical dilation <5 cm
Exclusion: vaginal bleeding, sexual intercourse, vaginal examination in previous 24 hrs.
Methods
fFN test done as per manufacturer instructions, test results within 24 hours (so not bedside version).
Assay results available 24-48 hours. There was a specific management protocol with fFN in place.
Results
Looked at fFN testing to determine effect on maternal transfers, cost of transfers, and local preterm delivery rates
and admissions to tertiary hospital.
151 patients admitted (98 to referral hospitals, 53 to tertiary).
No difference in maternal age or ethnicity of women, but fewer married and higher parity, in the pre- and post-fFN
periods.
There were significantly fewer patients admitted for PTL, length of stay and number of admissions per patient.
Fewer patients received tocolytics and the costs for admissions were lower.
There were no negative impacts on neonatal outcomes (admissions to NICU, NICU LOS, days of ventilator support
per NICU admitted patient).
Critical comments
Results available 24-48 hours – could have seen better impact with point of care version of test.
Due to design, can’t rule out fewer admissions due to fewer patients presenting with symptoms or changing practice
patterns in preterm labour management over time.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
54
© 2006, Institute of Health Economics
Giles W, Bisits A, Knox M, Madsen G, Smith R. The effect of fetal fibronectin testing on admissions to a tertiary maternal-fetal medicine
unit and cost savings. American Journal of Obstetrics & Gynecology 2000; 182(2):439-442.
Overall conclusion
Aim
“..to determine whether the introduction of routine fetal fibronectin bedside testing affected costs and transfer rates
from referral district hospital to a tertiary obstetric hospital, as well as direct admissions to a tertiary referral
hospital”
Setting
Australia. Referral district hospitals plus tertiary hospital.
Design
18 month prospective audit. No control comparison.
Methods
151 women who had the fFN test were studied.
Results
Critical comments
Cost savings in terms of transport costs for patients with a negative fFN result who weren’t admitted or transferred
were calculated.
90% of women with fFN negative were not transferred from referral hospital. This was defined as an unnecessary
transfer and cost savings calculated.
Looks only at reduced costs for fFN negative patients not admitted. (Assumes then they would ALL have been
admitted in the absence of fFN – doesn’t look for the reduction of admissions relative to management without fFN
testing available). This also doesn’t account for the potential for increased management costs of fFN positive
patients.
Very limited since no comparison group considered.
No neonatal outcomes were considered.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
55
© 2006, Institute of Health Economics
Grobman WA, Welshman EE, Calhoun EA, Ramsey PS. Fetal fibronectin results did not reduce medical resource use for women with preterm
uterine contractions. American Journal of Obstetrics and Gynecology 2004; 191, 235-40.
Overall conclusion
fFN testing had NO EFFECT on use of medical resources (LOS, use of tocolysis and corticosteroids, admissions.
Aim
“..to assess whether the availability of fetal fibronectin results for women who undergo evaluation for preterm labour would
affect physician resource use and health care costs.”
Setting
U.S. tertiary care facility
Design
RCT (non-blinded), if eligible for fFN testing all tested, randomly allocated as to whether physician knew test results.
Inclusion: 24-34 weeks gestation, singleton, uterine contractions (> 6 contractions per hour).
Exclusion: vaginal bleeding, cerclage, nonintact membranes, >3cm dilation, vaginal exam or sexual intercourse
within 24 hours. Also if already receiving treatment for preterm contractions.
Methods
The fFN test had not been used in labour and delivery prior to study, and was only available within study protocol during the
study.
Randomization occurred AFTER standardized evaluation, including fFN test followed by cervical exam.
Swabs analyzed right away for “results available” group; frozen then analyzed as a batch for the “no results” group.
fFN specimen collection as usual. No specific management protocol incorporated, see educational intervention below.
For “results available” group, results communicated to attending physician by resident.
Educational intervention to ensure attending physician “aware of test characteristics”:
•
Before study, letter mailed to admitting OBs
•
When results given, the following statement always made: “If a patient has a positive result, her
risk of delivery in next 7 days is approximately 13%, if she has a negative result her risk of
delivering is less than 1%. Your patient had a positive/negative test”.
All treatment decisions made by attending physician.
Repeat PTL was maintained in original groups. “Results available” group could order a second test if >7 days post.
Resource use data collected:
•
Number of hospital admissions
•
Number of hospital days
•
Hours on labour & delivery
•
Number of drug treatments
Medical billing provided costs of this resource use.
Patient related costs lost income, need to hire home assistance, lost time to usual activities, emotional impact.
Results
Sample size: 50 per group. Powered to detect a 20% cost difference.
No difference in gestational age, cervical dilation, socioeconomic backgrounds, pregnancy outcomes.
Predictive values of fFN similar to other studies: 8 positive results and 1 delivered within 1 week (PPV 12.5%), 5 delivered
before 37 weeks (PPV 62.5%). 91 women had negative result and 3 delivered within one week (NPV 96.7%), 17 were
delivered before 37 weeks (NPV 81.3%).
No difference in:
Time spent in labour and delivery for initial evaluation
Frequency of tocolysis or corticosteroids
Conversion to inpatient status
Hospital readmission post-discharge
Overall conclusion – no differences between the 2 groups with respect to any cost variable, including the primary outcome of
the study (total costs).
Critical comments
Conducted some sub-group analyses and did not find any indications of a difference by whether cervical dilation was
present, by time (did it take time for them to learn how to do fFN?), or by patients on public aid vs. others.
Are these attending physicians “too good” already? That is, are they already doing a great job of identifying the right
women to admit, and therefore not over prescribing etc?
•
Performed in a university hospital with 24-hour resident coverage and strict protocol for preterm labour
management.
No neonatal outcomes.
Educational intervention not sufficient to affect behaviour??
•
No assurance that the physicians understood the test.
What is the impact of freezing the samples on the integrity of the results?
Interesting quote: “..if physicians are reluctant to abandon interventions (such as bed rest) in the setting of a negative test
result and begin to use increased interventions after a positive test (which has a low predictive value), the number of medical
and social resources that are consumed actually will increase after the use of the test.”
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
56
© 2006, Institute of Health Economics
Lowe MP, Zimmerman B, Hansen W. Prospective randomized controlled trail of fetal fibronectin on preterm labour management in a
tertiary care centre. American Journal of Obstetrics and Gynecology 2004; 190(2): 358-62.
Overall conclusion
No difference in length of hospital stay and interventions. Some evidence of shorter length of stay in labour and
delivery for fFN group.
Aim
“..to investigate the effect of the rapid fFN on the length of hospital stay and the use of preterm labour interventions.”
Setting
U.S. tertiary care center.
Design
RCT
Women with signs and symptoms of PTL (uterine contractions and/or cervical change) examined at the labour and
delivery unit, or those transferred there and already receiving tocolytics.
Inclusion: GA between 23-34 weeks, > 16 years of age, cervical dilation <=3cm for primigravid women and <=4 cm
for multiparous women.
Exclusion: higher order (than twin) multifetal gestations, cerclage, preterm premature rupture of membranes, vaginal
bleeding.
Randomization stratified by GA.
Methods
fFN specimen collection usual as per manufacturer. Preterm labour management was left to the discretion of the
treating physician.
Women randomized to either: PTL management with fFN (as soon as criteria met) or PTL management without fFN.
That is, women who had had a cervical ultrasound, transvaginal ultrasound or intercourse within the previous 24
hours were enrolled, but fFN test delayed.
Results
97 women were included (46 with fFN, 51 without).
No difference in age, gravidity, parity, previous preterm birth, referring physician, multiple gestations, or GA at time
of test. No difference in cervical dilation or effacement.
No significant difference between groups in terms of: time spent in labour and delivery, admission to the antepartum
unit, length of stay on the antepartum ward, use of tocolysis, betamethasone and antibiotics.
No difference when the GA subgroups were analyzed.
However, they also used a Cox proportional hazard model, with an adjustment for previous preterm birth, cervical
dilation and GA which showed that women in the fFN group had a shorter length of stay in labour and delivery.
Critical comments
Only study to stratify by GA, but no power for GA subgroup analysis
No test to ensure the assumptions of the Cox proportional hazard model were met.
They do not report on how many women in each group had had drugs administered prior to referral which may
confound the results.
The authors suggest their sample size may have been too small to detect a smaller difference in length of stay. They
powered based upon almost a 50% reduction.
No neonatal outcomes.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
57
© 2006, Institute of Health Economics
Mozurkewich E, Naglie G, Krahn M, Hayashi R. Predicting preterm birth: a cost-effectiveness analysis. Am J Obstet Gynecol 2000;182(6):
1589-1598.
Overall conclusion
Risk prediction strategies, including fetal fibronectin testing, may offer cost savings compared with treating all
women with threatened PTL and may prevent similar numbers of cases of respiratory distress syndrome and neonatal
deaths.
Aim
To compare the cost-effectiveness of 9 strategies for the management of threatened PTL
Setting
University hospital in Michigan, United States
Design
Decision analytic model of costs, cases of respiratory distress and neonatal deaths.
Uses published meta-analyses data where possible, institutional data for costs.
Methods
9 different strategies are considered in the decision analysis: The strategies considered are
(1) treat all with tocolytics and corticosteroids,
(2) administer ‘traditional’ fFN test and in meantime treat all, discontinuing those who are negative,
(3) administer cervical ultrasound (transvaginal or transperineal) and treat only with abnormal cervical length
measurements,
(4) administer ‘rapid’ fFN and treat those who are positive,
(5) administer rapid fFN plus ultrasound and treat those with abnormal results on at least one,
(6) treat none.
They further consider an additional 3 strategies –
(7) treat all women with outpatient corticosteroids only,
(8) corticosteroids to all, tocolytics only to those with positive rapid fFN test, and (9) corticosteroids to all, tocolytics
only to those with abnormal cervical length.
Results
Critical comments
Compared with treating all women with threatened preterm labour, the authors conclude that risk assessment
strategies using either fFN or cervical length assessment offer cost savings and prevent cases of respiratory distress
syndrome and neonatal death. The addition of corticosteroids improves the clinical outcomes at lower costs.
Traditional fFN no longer relevant, the rapid test is the only one available.
The baseline comparator is “treat all” women – however, clinical management would not involve treating all women.
Clinical judgement would determine who to treat (admit to hospital etc). This study does not carefully consider the
entry criteria into the model.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
58
© 2006, Institute of Health Economics
Musaad SMA, Melson CL, Boswell DR. Assessment of the impact of introducing fetal fibronectin assay in the management of preterm labour at
Middlemore Hospital, New Zealand. Pathology 2005; 37(3):226-230.
Overall conclusion
Total expenditure on patient management did not go down with fFN testing.
Learning curve speculated – downward trend noted.
Aim
“To assess he impact of introducing the fFN assay on the diagnosis, length of hospital stay and cost of managing patients
presenting with symptoms of premature labour in our hospital”
Setting
New Zealand.
Design
Observational, before and after fFN introduction, matched
historical controls.
Methods
The first 30 fFN tested patients were recruited and followed
until delivery. “Management of preterm labour” compared with 30
matched historical controls. Indicators of this were:
cost savings (hospital days, medications and investigations)
length of stay
Rate of in utero transfers
Delivery at gestations < 34 weeks
Neonatal outcomes (admission to NICU, ventilation, death).
Inclusion criteria: GA 24-33 weeks, symptoms and signs of PTL, intact membranes, dilation < 3 cm, no cervical effacement
(cervix > 1 cm long), no frank vaginal bleeding.
Twin pregnancies included AND sexual intercourse in 24 hours prior did NOT exclude patients.
fFN protocol: samples collected as per manufacturer instructions, used Rapid fFN Tli system, results available less then 12
hours.
Previous PTL management strategy for patients 24-33 weeks gestation: start tocolysis, steroids, antibiotics. Many had an
obstetric ultrasound.
New protocol: if fFN positive admit and begin drug treatments, offered an ultrasound scan. If negative, discharge home with
advice and follow-up.
Cost data was taken from estimates submitted to Lab in support of introducing fFN testing (cost per day, cost of drugs,
ultrasound and standard tests). NO detail given on the source of the original cost data. It indicates that calculations were
made based on actual tests and interventions performed and the mean length of hospital stay.
Results
No difference in baseline characteristics of two groups (age, ethnicity,
%twin gestations, gestational age.)
Clinical variables: NO statistically significant difference in use of
ultrasound, antibiotics, in utero transfer or median hospital days.
Statistically significant difference in use of tocolysis and steroids.
“Trend” towards lower median hospital stay in fFN group. There was a
lot of variability in LOS.
Control group: Median 2 days, Range 1-11, Mean(sd) 2.7 +/- 2.3.
fFN: Median 1, range 1-6, Mean(sd) 1.97 +/-1.43.
Costs variables: The control group has significantly lower costs for tests and
interventions, but higher costs for hospital stay so that overall Total costs are not
significantly different.
However – there were 7 in the fFN group with long LOS due to other co-morbidities. The authors speculated this
was driving costs and re-did the testing excluding these patients. In that case, total cost is still not statistically
different but they indicate there is a ‘trend’ towards lower costs.
Critical comments
There was no difference in GA at delivery, GA < 34 weeks, preterm delivery, and NICU admissions or ventilations.
Protocol violations: Sexual intercourse should be an exclusion criterion.
(Can result in false positives). They had a 25% fFN positive rate,
which is high.
Very small numbers involved.
Issues around patient selection – these patients with co-morbidities
that would warrant hospitalization probably did not need fFN testing.
They do not report same information about the control group.
Unclear how control group was selected. It looks like it is all who are
hospitalized versus a selection who presented with PTL. You would
think some would be hospitalized and some sent home based on
clinical indications. They do not report number of admissions, but 29
of 30 get steroids, so it appears that all were admitted.
No neonatal outcomes
More recent report of NZ study in presentation at conference.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
59
© 2006, Institute of Health Economics
Nguyen TCQ, Toy EC, Baker B . The cost-effectiveness of fetal fibronectin testing in suspected preterm labor: a randomized trial
[abstract]. Obstetrics & Gynecology 2002; 99(4 Suppl):97S, 2002.
Overall conclusion
fFN testing associated with longer triage stays and higher hospital charges
Aim
“To compare the cost-effectiveness of vaginal fetal fibronectin testing versus observation in evaluating suspected
preterm labour.
Setting
U.S.
Design
RCT
Women between 24-35 weeks gestation with symptoms of PTL were recruited.
Exclusion: abdominal trauma, bleeding, nonreassuring fetal heart tracing, history of tocolysis in the current
pregnancy or recent digital exam or intercourse.
Methods
Protocol: prior to digital exam – a sterile speculum exam to assess for ruptured membranes AND fFN swab
completed.
If membranes intact and dilation < 3cm, women were randomized to either:
“traditional observation” - swab discarded, observed with serial digital examinations.
“fFN testing”
If fFN negative – discharge home.
If positive – further observation.
Results
Looked at time spend in triage and hospital charges.
77 women enrolled – 35 in traditional group. 42 to fFN.
Mean age, gravidity, parity, estimated gestational age, initial cervical dilation, and hospitalization and tocolysis rates
were similar.
Critical comments
fFN women spend more time in triage unit (3.3 versus 2.7 hours, p=0.05) and had higher hospital charges ($452 vs.
$299, p=0.03)
Abstract only so ability to interpret is limited.
This is not a cost-effectiveness study.
Why history of tocolysis in current pregnancy excluded? Presumably to get only those presenting for first time with
PTL.
What is their triage process? The triage times are statistically significantly different, but is a 0.3 hour (18 minute)
difference clinically meaningful?
Were the results due to the apparently in appropriate use of positive fFN result - because all patients with positive
results underwent further observation? Positive predictive value not good enough to merit this.
No neonatal outcomes.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
60
© 2006, Institute of Health Economics
Plaut MM, Smith W, Kennedy K. Fetal fibronectin: the impact of a rapid test on the treatment of women with preterm labour symptoms.
American Journal of Obstetrics and Gynecology 2003; 188(6): 1588-1593.
Overall conclusion
Overall, no difference in length of stay. For high risk women, there was a significant reduction in length of stay.
Aim
“..to evaluate whether knowledge of the results of this rapid [fFN] test would decrease unnecessary interventions for
patients with symptoms of preterm labour and perhaps to better identify that group of patients who might benefit
from aggressive therapy that would include tocolysis, corticosteroids, and transfer to a tertiary care facility.”
Setting
U.S., four community hospitals in Oregon/Washington (three used in analysis because one only enrolled one patient).
Design
RCT
Inclusion: present with symptoms of PTL 24 -34 weeks 6 days.
Exclusion: cervical manipulation (intercourse, vaginal examination or transvaginal ultrasound) within previous 24
hours, confirmed rupture of membranes, bleeding, dilation greater or equal to 3 cm, cervical cerclage, previous fFN
testing within 2 weeks)
Methods
fFN protocol: usual as per manufacturer. Available only as part of study protocol. Test available in 1-2 hours.
All women had the fFN test, and laboratory processed results – randomization took place at the laboratory where the
lab personnel were given random instructions to either inform physician of results or inform physician the patient
was in the no results group.
Treatment decisions at the physician’s discretion.
Educational intervention to inform physician’s about the test: ‘standardized’ educational materials distributed in
conferences, as written materials, and as posters in labour and delivery areas.
Study was powered for primary outcome of transport to tertiary care centres (n=500) but stopped prematurely for
lack of enrolment. Analysis thus focussed on secondary outcomes, focussing on length of stay.
Results
100 patients (108 swabs) – 8 patients were entered into the study twice because they re-presented with PTL
symptoms greater than 2 weeks after initial.
NO difference in characteristics of study groups (gestational age, gestational age at delivery, known risk factors, twin
gestations, fFN positive rate, parity).
Hospital stay: length of stay (including observation plus admission) not significantly shorter in fFN results known
group (6.8 versus 8.1, p=.35). However, for those observed more then six hours there was a statistically significant
decrease in LOS from 37.8 to 22.7 (p=.04)
Critical comments
Treatment of readmissions? They did not follow costs up to birth.
Educational intervention adequate? “One could argue that, because there was no required protocol for negative or
positive tests in the group for whom the result was known, we may have seen no difference in some outcomes simply
because physicians ignored the results of the test. Our study attempted to mimic the real world, where physicians are
educated about a test then try to incorporate it into their practice.”
Study stopped early. In the end, they did not have sufficient power even for secondary analyses. The confidence
intervals around estimates are large.
6 of 47 patients with negative fFN (results known) had aggressive therapy – this is indicative of potentially
inappropriate interpretation of fFN results.
For more difficult to diagnose patients (indicated by longer observation time), fFN testing did significantly reduce
LOS by 40%.
No neonatal outcomes.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
61
© 2006, Institute of Health Economics
Sullivan A, Hueppchen NA, Satin AJ. Cost effectiveness of bedside fetal fibronectin testing varies according to treatment algorithm.
Journal of Maternal-Fetal Medicine 2001; 10(6):380-384.
Overall conclusion
fFN testing on ALL patients presenting with PTL significantly increases costs; fFN
testing after decision to admit based on clinical criteria may reduce the cost.
Aim
Setting
Design
Methods
Results
Critical comments
“To determine the cost-effectiveness of implementing fetal fibronectin testing in
women with threatened preterm labour.”
U.S. hospital
Decision analysis model of cost outcomes (hospital admission
and fFN cost). Uses institutional data on threatened PTL, estimates
from the literature.
Three strategies are compared in the decision analytic
model:
1. fFN assay and examination
2. examination and fFN assay if decision to admit
3. examination only
With varying prevalence of threatened preterm labour and varying admission rates, fFN testing on all patients will
increase admissions and increase costs but fFN testing on only those who would be admitted will results in decreased
admissions and lower costs.
No inclusions of cost of follow-up protocol for negative fFN assay patients or repeat episodes of PTL.
Assumptions about rate of positive fFN in women who would be admitted by clinical criteria is not based on any
evidence.
No neonatal outcomes.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
62
© 2006, Institute of Health Economics
APPENDIX 3
Diagnosis codes used for initial selection of patients from AHW databases
O47003
O47103
O47903
O75801
O75802
O75803
O75804
O75809
O42011
O42013
O42019
O42021
O42023
O42029
False Lab <37 Comp Wks Gest/Antepart
FALSE LABOUR BEFORE 37
COMPLETED WEEKS OF GESTATION, ANTEPARTUM CONDITION OR
COMPLICATION
False Lab >=37 Comp Wks Gest/Antepart
FALSE LABOUR AT OR
AFTER 37 COMPLETED WEEKS OF GESTATION, ANTEPARTUM
CONDITION OR COMPLICATION
False Labour Unspec/Antepartum
FALSE LABOUR, UNSPECIFIED,
ANTEPARTUM CONDITION OR COMPLICATION
Preterm Labour W Del Delay Rx/Delivered
PRETERM LABOUR WITH
DELIVERY DELAYED BY THERAPY, DELIVERED, WITH OR WITHOUT
MENTION OF ANTEPARTUM CONDITION
Preterm Labour W Del Delay Rx/Del W Comp PRETERM LABOUR WITH
DELIVERY DELAYED BY THERAPY, DELIVERED, WITH MENTION OF
POSTPARTUM COMPLICATION
Preterm Labour W Del Delay Rx/Antepartum PRETERM LABOUR WITH
DELIVERY DELAYED BY THERAPY, ANTEPARTUM CONDITION OR
COMPLICATION
Preterm Labour W Del Delay Rx/Postpartum
PRETERM LABOUR WITH
DELIVERY DELAYED BY THERAPY, POSTPARTUM CONDITION OR
COMPLICATION
Preterm Labour W Del Delay Rx/Unspec
PRETERM LABOUR WITH
DELIVERY DELAYED BY THERAPY, UNSPECIFIED AS TO EPISODE OF
CARE, OR NOT APPLICABLE
Preterm Prm Lab 24 Hr/Delivered
PRETERM PREMATURE RUPTURE
OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS, DELIVERED,
WITH OR WITHOUT MENTION OF ANTEPARTUM CONDITION
Preterm Prm Lab 24 Hrs/Antepartum PRETERM PREMATURE RUPTURE
OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS,
ANTEPARTUM CONDITION OR COMPLICATION
Preterm Prm Lab 24 Hrs/Unspec
PRETERM PREMATURE RUPTURE
OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS,
UNSPECIFIED AS TO EPISODE OF CARE, OR NOT APPLICABLE
Full Term Prm Lab 24 Hrs/Deliv
FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS,
DELIVERED, WITH OR WITHOUT MENTION OF ANTEPARTUM
CONDITION
Full Term Prm Lab 24 Hrs/Antepartum FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS,
ANTEPARTUM CONDITION OR COMPLICATION
Full Term Prm Lab 24 Hrs/Unspec
FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS,
UNSPECIFIED AS TO EPISODE OF CARE, OR NOT APPLICABLE
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
63
© 2006, Institute of Health Economics
O42091
O42093
O42099
O42111
O42113
O42119
O42121
O42123
O42129
O42191
O42193
O42199
O42201
O42203
Prm Lab 24 Hrs Unspec/Deliv PREMATURE RUPTURE OF MEMBRANES,
ONSET OF LABOUR WITHIN 24 HOURS, UNSPECIFIED WHETHER
PRETERM OR FULL TERM, DELIVERED, WITH OR WITHOUT MENTION
OF ANTEPARTUM CONDITION
Prm Lab 24 Hours Unspec/Antepartum PREMATURE RUPTURE OF
MEMBRANES, ONSET OF LABOUR WITHIN 24 HOURS, UNSPECIFIED
WHETHER PRETERM OR FULL TERM, ANTEPARTUM CONDITION OR
COMPLICATION
Prm Lab 24 Hrs Nos/Unspec
PREMATURE RUPTURE OF MEMBRANES,
ONSET OF LABOUR WITHIN 24 HOURS, UNSPECIFIED WHETHER
PRETERM OR FULL TERM, UNSPECIFIED AS TO EPISODE OF CARE, OR
NOT APPLICABLE
Preterm Prm Lab Aft24 Hrs/Deliv
PRETERM PREMATURE RUPTURE
OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS, DELIVERED,
WITH OR WITHOUT MENTION OF ANTEPARTUM CONDITION
Preterm Prm Lab Aft 24 Hrs/Antepartum
PRETERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS,
ANTEPARTUM CONDITION OR COMPLICATION
Preterm Prm Lab Aft 24 Hrs/Unspec
PRETERM PREMATURE RUPTURE
OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS, UNSPECIFIED
AS TO EPISODE OF CARE, OR NOT APPLICABLE
Full Term Prm Lab Aft 24 Hrs/Deliv
FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS.
DELIVERED, WITH OR WITHOUT MENTION OF ANTEPARTUM
CONDITION
Full Term Prm Lab 24 Hrs/Anterpartum FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS,
ANTEPARTUM CONDITION OR COMPLICATION
Full Term Prm Lab Aft 24 Hrs/Unspec FULL TERM PREMATURE
RUPTURE OF MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS,
UNSPECIFIED AS TO EPISODE OF CARE, OR NOT APPLICABLE
Prm Lab Aft 24 Hours Unspec/Deliv
PREMATURE RUPTURE OF
MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS, UNSPECIFIED
WHETHER PRETERM OR FULL TERM, DELIVERED, WITH OR
WITHOUT MENTION OF ANTEPARTUM CONDITION
Prm Lab Aft 24 Hrs Unspec/Antepartum PREMATURE RUPTURE OF
MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS, UNSPECIFIED
WHETHER PRETERM OR FULL TERM, ANTEPARTUM CONDITION OR
COMPLICATION
Prm Lab Aft 24 Hrs Nos/Unspec
PREMATURE RUPTURE OF
MEMBRANES, ONSET OF LABOUR AFTER 24 HOURS, UNSPECIFIED
WHETHER PRETERM OR FULL TERM, UNSPECIFIED AS TO EPISODE
OF CARE, OR NOT APPLICABLE
Prom Labour Delayed By Therapy/Deliv
PREMATURE RUPTURE OF
MEMBRANES, LABOUR DELAYED BY THERAPY, DELIVERED, WITH
OR WITHOUT MENTION OF ANTEPARTUM CONDITION
Prom Labour Delayed By Therapy/Antepart
PREMATURE RUPTURE OF
MEMBRANES, LABOUR DELAYED BY THERAPY, ANTEPARTUM
CONDITION OR COMPLICATION
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
64
© 2006, Institute of Health Economics
O42209
O42909
O60001
Prom Lab Delayed By Therapy/Unspec PREMATURE RUPTURE OF
MEMBRANES, LABOUR DELAYED BY THERAPY, UNSPECIFIED AS TO
EPISODE OF CARE, OR NOT APPLICABLE
Prom Nos/Unspecified PREMATURE RUPTURE OF MEMBRANES,
UNSPECIFIED, UNSPECIFIED AS TO EPISODE OF CARE, OR NOT
APPLICABLE
ICD10CA
Preterm Delivery/Delivered
PRETERM DELIVERY,
DELIVERED, WITH OR WITHOUT MENTION OF ANTEPARTUM
CONDITION
CMG codes: 599 (Premature labour) or 619 (False Labour LOS < 3 Days (MNRH)
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
65
© 2006, Institute of Health Economics
APPENDIX 4
Fetal Fibronectin Health Technology Project
Data Requirements Summary
Data Year: 2004/2005
Methodology:
1. Identify all individuals in the inpatient database with a diagnosis of O42.XXX,
O47.XXX, O75.XXX or O60.001 or grouping to CMG 599 or CMG 619.
Exclude: Records with Responsibility for Payment <> “01” (Provincial
Government) and invalid PHNs.
2. Identify all individuals in the ambulatory care database with a diagnosis of
O42.XXX, O47.XXX, O75.XXX or O60.001
Exclude: Records with Responsibility for Payment <> “01” (Provincial
Government) and invalid PHNs.
3. For individuals identified, provide demographic information and information on
all inpatient, ambulatory care and physician services received during the first
encounter and following the first encounter until fiscal year end (as outlined
below).
1. Population Registry Database (March 31st, 2005)
Record Count: 17,855
•
•
•
•
•
•
Recipient Identifier (Randomly generated)
Recipient Age
Recipient Gender
Recipient Health Region of Residence
Recipient Postal Code (3 first digits only, will link to income level
information from Census)
Socioeconomic Status
2. Inpatient Database (2004/2005)
Record Count: 20,256
•
•
•
•
•
•
Recipient Identifier (Randomly generated)
Recipient Health Region of Residence at Service Date
Recipient Age at Service Date
Discharge Disposition
Admission Date
Discharge Date
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
66
© 2006, Institute of Health Economics
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Length of Stay
Facility Number
Transfer To
Transfer From
Admit Via Ambulance
Main Diagnosis
Other Diagnoses (24 occurrences)
Diagnosis Type
Main Intervention
Other Interventions (19 occurrences)
Average Length of Stay
Case Mix Group (CMG)
Resource Intensity Weight (RIW)
Main Patient Service
Service Transfers (3 occurrences)
Provider Service (8 occurrences)
Responsibility for Payment
Provider RHA
Number of Previous Term Deliveries
Number of Previous Pre-term Deliveries
Gestational Age
3. Ambulatory Care Database (2004/2005)
Record Count: 48,217
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Recipient Identifier (Randomly generated)
Recipient Health Region of Residence at Service Date
Discharge Disposition
Visit Start Date
Visit End Date
Facility Number
MIS Functional Centre
Transfer To
Transfer From
Admit Via Ambulance
Main Diagnosis
Other Diagnoses (9 occurrences)
Main Intervention
Other Interventions (9 occurrences)
ACCS Group
System Wide Relative Value
Provider Type (5 occurrences)
Doctor Service (5 occurrences)
Responsibility for Payment
Provider RHA
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
67
© 2006, Institute of Health Economics
4. FFS Physician Database
Record Count: 248,689
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Recipient Identifier (Randomly generated)
Recipient Health Region of Residence at Service Date
Service Start Date
Service End Date
Provider Identifier
Provider Discipline
Provider Specialty
Program Type
Facility Number
Functional Centre of Facility
Diagnosis (3 occurrences)
Service Code
Amount Paid
Service Health Region
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
68
© 2006, Institute of Health Economics
APPENDIX 5
Information on all air and combined air/ground ambulance transfers
Examining the entire dataset, 200 air or combined air/ground ambulance transfers can be
identified.
• For 57 trips in total (22 air, 35 combined), the woman’s gestational week was
greater than 35 therefore fFN is not applicable.
• For an additional 55 trips, there is a gestational age noted in the database:
o 54 of these are for preterm birth, and are therefore unavoidable
o 1 is for premature rupture of membranes, therefore fFN is not applicable
• For 88 trips (37 air, 51 combined), the gestational age is missing. These were
examined in detail to determine the reason for the trip and also to infer the
gestational age at the time of the trip based on the gestational age at birth.
Examining the 37 air ambulance trips:
• 3 were not for maternal diagnoses, 7 were maternal diagnoses but not PTL or fFN
not applicable, 10 were for premature rupture of membranes, 2 were for diagnosis
of PTL but gestational age greater than 37 weeks, 1 was for diagnosis of PTL and
gestational age was < 37 but there were other complications, 3 were for diagnosis
of PTL delayed by therapy but there were other complications. Thus, 26 of the
trips are not avoidable.
• 4 trips were for diagnosis of PTL and 7 trips were for diagnosis of PTL delayed
by therapy. Thus there are 11 potentially avoidable air transfers depending on the
gestational age – these were examined in more detail.
From RHA
To RHA
Birth
outcome
GA at Birth
RHA at
Birth
Calculated
GA at
transfer
34.4
36.3
Aspen
Community
Capital
Northern
Lights
32 days later
5 days later
39 weeks
37 weeks
Capital*
Northern
Lights
Peace
Aspen
Capital
Capital
37 days later
48 days later
38 weeks
42 weeks
Capital*
Capital*
32.7
35.1
David
Thompson
Aspen
Capital
32 days later
34 weeks
29.4
Capital
40 weeks
24
Potentially
Peace
Aspen
Capital
Capital
105 days
later
64 days later
32 days later
David
Thompson
Peace
Potentially
No, from
community,
GA>35
Potentially
Potentially
GA = 35
Potentially
40 weeks
40 weeks
Peace
Aspen
30.9
35.4
Aspen
Capital
31 days later
37 weeks
Capital*
35.6
Peace
Northern
Lights
Capital
Capital
33 days later
49 days later
39 weeks
38 weeks
Peace
Northern
Lights
34.3
31
Potentially
Potentially
(GA = 35)
Not, original
GA>35
Potentially
Potentially
Avoidable?
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
69
© 2006, Institute of Health Economics
* Note that for these three transfers to Capital, the birth ultimately occurred in Capital as well. For
only one was there an additional ambulance transfer, the other two either remained in Capital for the
remainder of their pregnancies or returned to Capital via personal transport for their births.
Of the 11 original air transfers examined in more detail, 2 can be removed as the
gestational age at the time of the transfer was calculated (using the GA at birth) to be
greater than 35 weeks.
¾ 9 avoidable air transfers, $31, 896 in avoidable costs
Examining the 51 combined air/ground transfers:
•
•
9 were for maternal diagnoses but not PTL or fFN not applicable, 6 were for
premature rupture of membranes, 8 were for PTL > 37 weeks, 3 for PTL with
complications, 4 for PTL delayed by therapy with complications. Thus 30 of
these trips are not avoidable.
10 were for PTL, 11 for PTL delayed by therapy. Thus, there are 21 potentially
avoidable trips to examine more closely.
From RHA
To RHA
Birth
outcome
GA at Birth
RHA at
Birth
Community
Northern
Lights
Northern
Lights
Capital
Calgary
Capital
Northern
Lights
Calgary
Capital
Unknown
Unknown
Unknown
98 days later
41 weeks
27
37 days later
This trip
40 days later
9 days later
40
34
29
37
This trip
34 days later
32
29
Capital
Unknown
Unknown
Northern
Lights
Capital*
Calgary
Capital*
Northern
Lights
Calgary
Northern
Lights
Unknown
Northern
Lights
Capital
Unknown
Unknown
Unknown
Unknown
40 days later
28
Community
Peace
Palliser
Aspen
Community
Palliser
Northern
Lights
Peace
Community
Northern
Lights
Northern
Lights
Northern
Lights
Northern
Lights
?
Peace
Northern
Lights
?
Calculated
GA at
transfer
Unknown
35
34
33
35
32
24
Unknown
Avoidable?
No, from
community
No, from
community
Potentially
No, birth
Potentially
No, from
community
No, birth
Potentially
32
Capital
Northern
Lights
Unknown (as of 35 weeks later)
Potentially,
can not
confirm GA
No, from
community
Potentially
Capital
24 days later
37
Capital*
34
No, assume
too early GA
Potentially
Capital
29 days later
38
Capital*
34
Potentially
Capital
Capital
Capital
25 days later
87 days later
26 days later
37
40
37
32
28
33
Potentially
Potentially
Potentially
Calgary
68 days later
41
Peace
Peace
Northern
Lights
Calgary*
32
Potentially
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
70
© 2006, Institute of Health Economics
From RHA
To RHA
Birth
outcome
GA at Birth
?
Capital
Not in data 23 weeks later
Aspen
Capital
11 days later
35
RHA at
Birth
Capital*
Calculated
GA at
transfer
33
Aspen
Capital
18 days later 37
Capital*
34
*For 7 of these transfers, they ultimately gave birth in Capital or Calgary as well.
Avoidable?
No, assume
too early GA
Potentially,
birth within
14 days
Potentially
Of the 21 transfers examined in more detail, 7 can be removed as they either resulted in
birth, were from the community or the gestational age at the time of transfer was
calculated to be > 35 weeks.
Of the remaining, 11 are potentially avoidable. Two others are considered potentially
avoidable as well – for one the GA can’t be confirmed, and for the other birth did occur
14 days later so this person may have tested positive for fFN and been transferred
anyways.
¾ Between 11 and 13 combination ambulance trips are potentially avoidable,
between $45,595 - $53,885.
In total, there are between $77,491 and $87,781 in potential savings due to
avoidable air or air/ground combined transfers.
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
71
© 2006, Institute of Health Economics
APPENDIX 6
Detailed Cost Calculations
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
72
© 2006, Institute of Health Economics
Detailed Cost Calculations - One Year Horizon
NUMBER OF ADMISSIONS
Level II
Level I
Level II and III
CHR
Capital
Other region All regions
PTL
125
136
60
73
PTL delay tx
69
165
31
15
PTB
390
328
101
27
584
629
192
115
394
280
846
1520
Scenario 2: fFN testing in Level II and III Hospitals
2a
2b
Costs of implementing fFN
Fixed costs
Variable costs
$2,668
$136,664
$2,668
$136,664
Costs avoided
Ground transfers avoided
Air transfers avoided
LOS reductions
$500
$0
$454,736
$500
$0
$419,428
Net cost avoidance
$315,904
$280,596
In 2a, assume full potential LOS reduction for Level II hospitals outside Capital/Calgary
In 2b, assume 50% of admissions at Level II hospitals outside Capital/Calgary.
could see a LOS reduction b/c average LOS close to 1
Scenario 1: fFN testing in all facilities
1a
1b
1c
1d
1e
1f
Costs of implementing fFN
Fixed costs
Variable costs
$162,748
$147,850
$162,748
$147,850
$162,748 $162,748 $162,748
$147,850 $147,850 $147,850
$162,748
$147,850
Costs avoided
Ground transfers avoided
Air transfers avoided
LOS reductions
$21,000
$77,500
$419,428
$42,000
$88,000
$419,428
$21,000 $21,000 $42,000
$77,500 $77,500 $88,000
$453,572 $487,716 $453,572
$42,000
$88,000
$487,716
Net cost avoidance
$207,330
$238,830
$241,474 $275,618 $272,974
$307,118
Change from Scenario 2b
-$73,266
-$41,766
-$39,122
-$4,978
-$7,622
$26,522
1a assume lower amount for both transfers, no additional LOS reductions to Scenario 1b
1b assumes higher amount for both transfers, no additional LOS reductions to Scenario 1b
1c assumes lower amount for both transfers, additional LOS reduction on 50% of Level I admissions
1d assumes lower amount for both transfers, additional LOS reductions on all Level I admissions
1e assumes higher amount for both transfers, additional LOS reductions on 50% of Level 1 admissions
1f assumers higher amount for both transfers, additional LOS reductions on all Level I admissions
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
73
© 2006, Institute of Health Economics
Year 2: Scenario 1 (fFN testing in all facilities), no fixed costs in Year 2
1a
1b
1c
1d
1e
1f
Costs of implementing fFN
Fixed costs
Variable costs
$147,850
$147,850
$147,850 $147,850 $147,850
$147,850
Costs savings
Ground transfers avoided
Air transfers avoided
LOS reductions
$21,000
$77,500
$419,428
$42,000
$88,000
$419,428
$21,000 $21,000 $42,000
$77,500 $77,500 $88,000
$453,572 $487,716 $453,572
$42,000
$88,000
$487,716
$370,078
$401,578
$120,982
$404,222 $438,366 $435,722
$123,626 $157,770 $155,126
$469,866
$189,270
Net costs savings
Change from Scenario 2b:
1a assume lower amount for both transfers, no additional LOS reductions to Scenario 1b
1b assumes higher amount for both transfers, no additional LOS reductions to Scenario 1b
1c assumes lower amount for both transfers, additional LOS reduction on 50% of Level I admissions
1d assumes lower amount for both transfers, additional LOS reductions on all Level I admissions
1e assumes higher amount for both transfers, additional LOS reductions on 50% of Level 1 admissions
1f assumers higher amount for both transfers, additional LOS reductions on all Level I admissions
Scenario 1 (fFN testing in all facilities)
Per year, assuming fixed costs spread over two years
1a
1b
1c
1d
1e
1f
Costs of implementing fFN
Fixed costs
Variable costs
$81,374
$147,850
$81,374
$147,850
$81,374 $81,374 $81,374
$147,850 $147,850 $147,850
$81,374
$147,850
Costs savings
Ground transfers avoided
Air transfers avoided
LOS reductions
$21,000
$77,500
$419,428
$42,000
$88,000
$419,428
$21,000 $21,000 $42,000
$77,500 $77,500 $88,000
$453,572 $487,716 $453,572
$42,000
$88,000
$487,716
Net costs savings
Change from Scenario 2b:
$288,704
$8,108
$320,204
$39,608
$322,848 $356,992 $354,348
$42,252 $76,396 $73,752
$388,492
$107,896
1a assume lower amount for both transfers, no additional LOS reductions to Scenario 1b
1b assumes higher amount for both transfers, no additional LOS reductions to Scenario 1b
1c assumes lower amount for both transfers, additional LOS reduction on 50% of Level I admissions
1d assumes lower amount for both transfers, additional LOS reductions on all Level I admissions
1e assumes higher amount for both transfers, additional LOS reductions on 50% of Level 1 admissions
1f assumers higher amount for both transfers, additional LOS reductions on all Level I admissions
Fetal Fibronectin Testing for the Diagnosis of Preterm Labour: A Review of the Economics Literature and Cost Analysis for Alberta
74
© 2006, Institute of Health Economics